CN115089179B

CN115089179B - Psychological emotion insight analysis method and system

Info

Publication number: CN115089179B
Application number: CN202210677204.3A
Authority: CN
Inventors: 白雪扬
Original assignee: Beijing Xueyang Technology Co ltd
Current assignee: Beijing Xueyang Technology Co ltd
Priority date: 2022-06-15
Filing date: 2022-06-15
Publication date: 2025-08-19
Anticipated expiration: 2042-06-15
Also published as: CN115089179A

Abstract

The invention provides a psychological emotion insight analysis method which comprises the steps of collecting blood light reflection signals, pulse wave signals and electrocardiosignals of a user through wearing equipment, preprocessing the blood light reflection signals, the pulse wave signals and the electrocardiosignals to obtain human physiological signals, calculating and analyzing the human physiological signal values to obtain multiple physiological indexes of the user, wherein the physiological indexes comprise heart rate, blood oxygen saturation, atrial fibrillation and blood pressure, and identifying and judging psychological emotion states of the user through a cosine similarity algorithm to obtain emotion judgment results and displaying the emotion judgment results on a mobile terminal in real time. The invention realizes the real-time monitoring and analysis of the psychological emotion of the user through daily common wearing equipment.

Description

Psychological emotion insight analysis method and system

Technical Field

The invention relates to the field of data acquisition and analysis of wearable equipment, in particular to a psychological emotion insight analysis method and system.

Background

Emotion is a state that integrates human feeling, thought and behavior, and plays an important role in human-to-human communication. Emotion is a state that integrates the feeling, thought and behavior of a person, and includes psychological reactions of a person to external or self-stimuli, including physiological reactions accompanying such psychological reactions. In daily work and life of people, the emotion function is ubiquitous. In medical care, if the emotional state of a patient, particularly a patient with dysexpressia, can be known, different care measures can be taken according to the emotion of the patient, and the care amount can be increased. In the product development process, if the emotional state of the user in the product use process can be identified, the user experience is known, the product function can be improved, and the product more suitable for the user requirement is designed. In various human-computer interaction systems, human-machine interaction becomes more friendly and natural if the system can recognize the emotional state of a person. Therefore, the analysis and recognition of emotion is an important interdisciplinary research topic in the fields of neuroscience, psychology, cognitive science, computer science, artificial intelligence and the like.

The existing emotion recognition technology mainly comprises a text analysis technology and a micro-expression recognition technology, wherein the text analysis technology is used for calculating emotion indexes of a user based on a sentence-level emotion analysis method by acquiring text contents of all utterances of the user, classifying psychological emotions of the user based on a cosine similarity method and finally displaying psychological evaluation and early warning results of the user. The micro-expression recognition technology recognizes micro-expressions of visitors in the psychological consultation process by acquiring the psychological consultation video of the user, acquires all time nodes and marks the time nodes, and judges psychological emotional conditions of the user. But two emotion recognition techniques have the following main problems:

1. the measuring process is complicated, namely, professional equipment is required to be worn for recording the expression, the behavior and the like of the user, and professional staff is also required to cooperate for marking and reading the records.

2. The real-time monitoring cannot be performed, and the conventional psychological emotion recognition method is limited by external equipment and the like and cannot be performed.

Disclosure of Invention

The invention provides a psychological emotion insight analysis method which is used for realizing real-time monitoring and analysis of psychological emotions of a user through daily common wearing equipment.

The invention provides a psychological emotion insight analysis method, which comprises the following steps:

collecting blood light reflection signals, pulse wave signals and electrocardiosignals of a user through a wearing device, and preprocessing the blood light reflection signals, the pulse wave signals and the electrocardiosignals to obtain human physiological signals;

calculating and analyzing the human physiological signal values to obtain a plurality of physiological indexes of a user, wherein the physiological indexes comprise heart rate, blood oxygen saturation, atrial fibrillation and blood pressure;

and identifying and judging the psychological and emotional states of the user through a cosine similarity algorithm by utilizing a plurality of physiological indexes of the user to obtain emotion judgment results, and displaying the emotion judgment results in real time on the mobile terminal.

Preferably, the preprocessing of the blood light reflection signal, the pulse wave signal and the electrocardiosignal includes:

Filtering baseline drift and ambient light noise caused by power frequency interference, breath jitter and the like in blood light reflection signals, pulse waves and electrocardiosignals through digital filtering;

after digital filtering, removing blood flow sounds in blood light reflection signals, pulse waves and electrocardiosignals through adaptive filtering;

and finally, carrying out mean value filtering and spike filtering on the blood light reflection signals, the pulse waves and the electrocardiosignals to eliminate coarse errors.

Preferably, the calculating and analyzing the human physiological signal value to obtain a plurality of physiological indexes of the user includes:

determining the heart rate and atrial fibrillation of the user through calculation according to the electrocardiosignals in the human physiological signals;

According to the blood light reflection signals in the human physiological signals, determining the blood oxygen saturation of the user through calculation;

and determining the blood pressure of the user through calculation according to the pulse wave signals in the human physiological signals.

Preferably, the identifying and judging the psychological emotion state of the user through the cosine similarity algorithm by using a plurality of physiological indexes of the user includes:

Acquiring a plurality of physiological indexes obtained by analyzing a blood light reflection signal, a pulse wave signal and an electrocardiosignal at the current moment so as to be acquired on the user;

Determining historical average values of all physiological indexes of the user according to the historical physiological index data under the user account stored in the cloud server, and establishing a normal physiological index comparison group by utilizing a plurality of historical average values;

determining a relative deviation value between each physiological index corresponding to the current moment and a historical average value of a corresponding physiological index in the normal physiological index control group;

If the relative deviation value between a certain physiological index and the historical average value of the corresponding physiological index in the normal physiological index control group is larger than a preset deviation threshold value, judging that the physiological index of the user is abnormal;

When determining that the physiological index of the user is abnormal, establishing a first calculation group according to a plurality of relative deviation values corresponding to the current physiological indexes of the user;

Cosine similarity calculation is carried out on the first calculation group and a plurality of second calculation groups stored in the cloud, so that similarity values between the first calculation group and the plurality of second calculation groups stored in the cloud are obtained;

and outputting the emotion nouns bound corresponding to the second calculation group corresponding to the highest similarity value as emotion judgment results.

Preferably, the emotion intensity of the user needs to be calculated while outputting the emotion judgment result by the following method:

determining a plurality of relative deviation values in the second calculation group corresponding to the highest similarity value;

calculating second relative deviation values between the plurality of relative deviation values in the first calculation group and the corresponding plurality of relative deviation values in the second calculation group respectively;

Carrying out average value calculation on a plurality of second relative deviation values to obtain the emotion intensity degree of the user;

When the emotion intensity of the user is larger than a preset fluctuation threshold value, the mobile terminal reminds the user to pay attention to adjust the emotion of the user.

Preferably, the second calculation group is determined by:

Determining a plurality of corresponding relative deviation values between a plurality of physiological indexes of a volunteer user under anger emotion and historical average values of corresponding physiological indexes in a normal physiological index comparison group of the volunteer user through volunteer data research;

Establishing a third calculation group according to a plurality of relative deviation values corresponding to the current physiological indexes of the volunteer user;

And determining a plurality of third computing groups which are correspondingly established by the plurality of volunteer users under the anger emotion, and carrying out corresponding mean value computation on the data in the plurality of third computing groups to obtain a second computing group corresponding to the anger emotion.

Preferably, the identifying and judging the psychological emotional state of the user through the cosine similarity algorithm by using the multiple physiological indexes of the user further includes:

Dividing a plurality of physiological indexes obtained by analyzing blood light reflection signals, pulse wave signals and electrocardiosignals acquired at the same moment into an index reference group;

setting an upper boundary value and a lower boundary value for each physiological index in advance, and comparing each current physiological index of a user with the corresponding preset upper boundary value and lower boundary value respectively;

If a certain physiological index in a certain index reference group is higher than the upper boundary value or lower than the lower boundary value corresponding to the preset physiological index, determining that the index reference group is abnormal, judging whether the index reference group corresponding to a plurality of subsequent moments is abnormal or not, and recording the occurrence times of the abnormality;

When the continuous times are larger than the upper limit of the preset times, determining that the user has emotion fluctuation, and continuously extracting a plurality of index reference groups from the index reference group with the abnormality for the first time;

classifying the multiple physiological indexes in the extracted multiple index reference groups according to item categories, sequentially arranging the multiple physiological indexes, establishing trend change curves of each physiological index, finally obtaining trend change curves corresponding to the multiple physiological indexes, and establishing a comparison data group by utilizing the trend change curves corresponding to the multiple physiological indexes;

Taking the corresponding comparison data templates of the plurality of psychological emotions from the cloud, respectively carrying out cosine similarity calculation on the comparison data templates of each psychological emotion and the comparison data sets, and determining the matching degree between the comparison data sets and the corresponding comparison data templates of each psychological emotion;

And outputting the psychological emotion corresponding to the comparison data template with the largest matching degree as an emotion judgment result.

Preferably, the upper boundary value and the lower boundary value are determined by:

obtaining a first fluctuation range of a certain physiological index of a human body in a normal emotion state through data research of volunteers or user group feedback;

Acquiring a second fluctuation range of a certain physiological index of the current user in a normal emotion state during the use of the wearable device;

Determining a first fluctuation median according to the first fluctuation range, wherein the first fluctuation median is the average value of the maximum value and the minimum value of the first fluctuation range;

Determining a second fluctuation median according to the second fluctuation range, wherein the second fluctuation median is the average value of the maximum value and the minimum value of the second fluctuation range;

Determining an absolute difference value of the first fluctuation median value and the second fluctuation median value, and multiplying the absolute difference value by a preset proportionality coefficient to obtain a corrected value;

Taking the maximum value in the second fluctuation range as a first upper boundary value, and adding the correction value to the first upper boundary value to obtain an upper boundary value;

And taking the minimum value in the second fluctuation range as a first lower boundary value, and adding the correction value to the first lower boundary value to obtain a lower boundary value.

Preferably, the step of performing cosine similarity calculation on the comparison data templates of each psychological emotion and the comparison data sets, and determining the matching degree between the comparison data sets and the comparison data templates corresponding to each psychological emotion includes:

selecting a certain psychological emotion comparison data template and carrying out similarity calculation on the template and the comparison data set;

when similarity calculation is performed, cosine similarity calculation is performed on all trend change curves in the comparison data set and trend change template curves of corresponding physiological indexes in the comparison data template respectively to obtain a plurality of first-class similarities, and each first-class similarity is bound with the corresponding physiological index item respectively, wherein the cosine similarity calculation process comprises the following steps:

determining two curves of the trend change curve and the trend change template curve;

Cutting the trend change template curve to obtain a first curve according to the length of the trend change curve;

Respectively carrying out equidistant quantization on the trend change curve and the second curve to obtain a plurality of sequentially arranged quantized data points;

Determining the slopes of a plurality of quantized data points on the trend change curve, and arranging the slopes of the plurality of quantized data points on the trend change curve to generate a first calculation sequence;

determining the slope of a plurality of quantized data points on the second curve, and arranging the slope of the plurality of quantized data points on the second curve to generate a second calculation sequence;

calculating by using the first calculation sequence and the second calculation sequence through a cosine similarity calculation formula to obtain the similarity between the trend change curve and the second curve as a first type of similarity;

And according to the weight coefficient preset for each corresponding physiological index in the comparison data template of the psychological emotion, carrying out weighted calculation based on a plurality of the first-class similarities to obtain the matching degree between the comparison data set and the comparison data template corresponding to the psychological emotion.

In order to achieve the above object, an embodiment of the present invention further provides a psychological emotion insight analysis system, including:

The signal acquisition processing module is used for acquiring blood light reflection signals, pulse wave signals and electrocardiosignals of a user through the wearing equipment, and preprocessing the blood light reflection signals, the pulse wave signals and the electrocardiosignals to obtain human physiological signals;

the signal analysis module is used for calculating and analyzing the human physiological signal value to obtain a plurality of physiological indexes of a user, wherein the physiological indexes comprise heart rate, blood oxygen saturation, atrial fibrillation and blood pressure;

the emotion judging module is used for identifying and judging the psychological emotion state of the user through a cosine similarity algorithm by utilizing a plurality of physiological indexes of the user to obtain an emotion judging result and displaying the emotion judging result in real time on the mobile terminal;

The identifying and judging the psychological emotion state of the user through the cosine similarity algorithm by utilizing a plurality of physiological indexes of the user comprises the following steps:

outputting the emotion nouns bound corresponding to the second calculation group corresponding to the highest similarity value as emotion judgment results;

The identifying and judging the psychological emotion state of the user through the cosine similarity algorithm by utilizing the multiple physiological indexes of the user further comprises the following steps:

outputting psychological emotion corresponding to the comparison data template with the largest matching degree as an emotion judgment result;

The step of calculating the cosine similarity between the comparison data templates of each psychological emotion and the comparison data sets, and the step of determining the matching degree between the comparison data sets and the corresponding comparison data templates of each psychological emotion comprises the following steps:

The technical scheme of the invention is further described in detail through the drawings and the embodiments.

Drawings

The accompanying drawings are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate the invention and together with the embodiments of the invention, serve to explain the invention. In the drawings:

FIG. 1 is a flow chart of a method for analyzing psychological emotion insight in an embodiment of the present invention;

FIG. 2 is a flow chart of a method for preprocessing a signal according to an embodiment of the present invention;

Fig. 3 is a schematic structural diagram of a psychological emotion insight analysis system in an embodiment of the present invention.

Detailed Description

The preferred embodiments of the present invention will be described below with reference to the accompanying drawings, it being understood that the preferred embodiments described herein are for illustration and explanation of the present invention only, and are not intended to limit the present invention.

The embodiment of the invention provides a psychological emotion insight analysis method, which comprises the following steps:

step S1, collecting blood light reflection signals, pulse wave signals and electrocardiosignals of a user through wearable equipment, and preprocessing the blood light reflection signals, the pulse wave signals and the electrocardiosignals to obtain human physiological signals;

S2, calculating and analyzing the human physiological signal values to obtain a plurality of physiological indexes of a user, wherein the physiological indexes comprise heart rate, blood oxygen saturation, atrial fibrillation and blood pressure;

And S3, identifying and judging the psychological and emotional states of the user by using a plurality of physiological indexes of the user through a cosine similarity algorithm to obtain emotion judgment results, and displaying the emotion judgment results in real time on the mobile terminal.

The technical scheme has the working principle and beneficial effects that daily wearable equipment such as a watch and a wrist band are used for collecting blood light reflection signals, pulse wave signals and electrocardiosignals of a user, the blood light reflection signals, the pulse wave signals and the electrocardiosignals are subjected to filtering and denoising pretreatment to obtain human physiological signals, then the human physiological signal values are calculated and analyzed to obtain multiple physiological indexes of the user, wherein the physiological indexes comprise heart rate, blood oxygen saturation, atrial fibrillation, blood pressure and the like, and finally the psychological emotion states of the user are identified and judged through a cosine similarity algorithm to obtain emotion judgment results and are displayed on a mobile terminal in real time. Thereby realizing continuous and uninterrupted psychological emotion insight analysis of the user.

In a preferred embodiment, preprocessing the blood light reflection signal, the pulse wave signal and the electrocardiographic signal includes:

step S21, filtering baseline drift and ambient light noise caused by power frequency interference, respiratory jitter and the like in blood light reflection signals, pulse waves and electrocardiosignals through digital filtering;

step S22, removing blood flow sounds in blood light reflection signals, pulse waves and electrocardiosignals through self-adaptive filtering after digital filtering;

And S23, finally, carrying out mean value filtering and spike filtering on the blood light reflection signals, the pulse waves and the electrocardiosignals to eliminate coarse errors.

The working principle and the beneficial effects of the technical scheme are that baseline drift and ambient light noise caused by power frequency interference, respiratory jitter and the like in blood light reflection signals, pulse waves and electrocardiosignals are removed through a digital filtering mode, blood flow sounds in the blood light reflection signals, the pulse waves and the electrocardiosignals are removed through self-adaptive filtering, coarse errors in the blood light reflection signals, the pulse waves and the electrocardiosignals are removed through mean filtering and spike filtering, and the filtering and noise removing processing of acquired signals is realized, so that more accurate signal data are obtained.

In a preferred embodiment, the calculating and analyzing the human physiological signal value to obtain a plurality of physiological indexes of the user includes:

determining heart rate and atrial fibrillation of a user through calculation according to electrocardiosignals in human physiological signals;

according to blood light reflection signals in the human physiological signals, determining the blood oxygen saturation of the user through calculation;

The technical scheme has the working principle and beneficial effects that the heart rate and atrial fibrillation of a user are determined through calculation by utilizing the electrocardiosignals in the physiological signals of the human body, the blood optical reflection signals in the physiological signals of the human body are utilized to determine the blood oxygen saturation of the user through calculation, the pulse wave signals in the physiological signals of the human body are utilized to determine the blood pressure of the user through calculation, and the conversion from the acquired basic signal wave data to the available index data is realized.

In a preferred embodiment, the identifying and judging the psychological emotional state of the user through the cosine similarity algorithm by utilizing a plurality of physiological indexes of the user comprises:

determining a relative deviation value between each physiological index corresponding to the current moment and a historical average value of the corresponding physiological index in the normal physiological index control group;

if the relative deviation value between the historical average value of a certain physiological index and the corresponding physiological index in the normal physiological index comparison group is larger than a preset deviation threshold value, judging that the physiological index of the user is abnormal;

The technical scheme has the working principle and beneficial effects that a plurality of physiological indexes are obtained by acquiring a blood light reflection signal, a pulse wave signal and an electrocardiosignal analysis from a user at the current moment, the historical average value of each physiological index of the user is determined according to the historical physiological index data stored in a cloud server under the user account, the specific values of each physiological index of the user under the daily condition are reflected through the historical average value (the emotion of the user cannot be analyzed by referring to a certain fixed historical physiological index template because of different physique between people, so that the historical physiological index of the user is very important as a comparison standard) and a normal physiological index comparison group is established by utilizing the plurality of historical average values, the relative deviation value between each physiological index corresponding to the current moment and the historical average value of the corresponding physiological index in the normal physiological index comparison group is determined, wherein the relative deviation value is the percentage between the physiological index of a certain item corresponding to the current moment and the average value of the corresponding physiological index of the corresponding item, if the difference between the physiological index of the certain item and the corresponding item of the normal physiological index is different from the historical average value of the daily condition, the relative deviation value is larger than the first physiological index of the user, the relative deviation between the current physiological index is determined when the first physiological index is larger than the corresponding physiological index of the current physiological index is calculated to be present, the plurality of data in the first calculation group can reflect the deviation condition between each physiological index of the user in the current period and each physiological index in daily life, and the cosine similarity calculation is carried out on the first calculation group and a plurality of second calculation groups stored in the cloud respectively, wherein the calculation formula is as follows:

Wherein Similar represents a similarity value, A _i represents an ith value in a first computing group, B _i represents an ith value in a second computing group, n represents the number of values in the first computing group or the second computing group, the second computing group can reflect the overall deviation situation between various physiological indexes of most users under a certain emotion and various physiological indexes in daily life, the similarity value between the first computing group and the cloud-stored second computing groups is obtained, and the emotion nouns corresponding to the second computing group corresponding to the highest similarity value are output as emotion judgment results, so that the comparison between the current physiological indexes and the daily physiological indexes of the users is realized, the deviation situation of the various physiological indexes of the users is reflected through the relative deviation, the similarity comparison calculation is carried out through the deviation situation templates of the common physiological indexes of the vast second computing group under a certain emotion (such as anger) of the users, the emotion psychological condition of the users is determined, and the emotion psychological emotion of the users is recognized by determining that the emotion of the users corresponding to the emotion situation of the most similar deviation situation template of the current physiological indexes of the users is the current.

In a preferred embodiment, the emotion intensity level of the user needs to be calculated while outputting the emotion judgment result by the following method:

Determining a plurality of relative deviation values in a second calculation group corresponding to the highest similarity value;

The working principle and the beneficial effects of the technical scheme are that when the emotion judgment result is output, a plurality of relative deviation values in a second calculation group corresponding to the highest similarity value are determined; calculating a second relative deviation value between a plurality of relative deviation values in the first calculation group and a plurality of corresponding relative deviation values in the second calculation group respectively, reflecting the deviation between the emotion intensity degree of a user under the current emotion and the normal emotion intensity degree through the second relative deviation, wherein the greater the deviation is, the more intense the emotion of the user, for example, a certain item of relative deviation value in the first calculation group is 10 percent to indicate that the heart beat speed of the user is 10 percent faster than daily, a relative deviation value of a corresponding item in the second calculation group is 20 percent to indicate that the heart beat speed of most users under the emotion is 20 percent faster than daily, the second relative deviation value is 10 percent minus 20 percent and then divided by 20 percent to be one half of negative, the higher the relative deviation value corresponding to a certain item of physiological index in the first calculation group is, the physiological index of the represented user changes more strongly relative to most users, the physiological index in this aspect is more intense because the physiological index of the user changes correspondingly, the relative intensity degree is not fixed by the relative items, the relative intensity degree is not fixed, the relative physiological index is calculated to be more than the average intensity degree of the second relative physiological index, the relative physiological index is more than the average intensity degree is calculated, and the relative physiological index is more than the average relative intensity degree is calculated, and the relative physiological index is more than the average degree is calculated relative intensity degree is more than the average degree, the mobile terminal reminds the user to pay attention to adjust the emotion of the user.

In a preferred embodiment, the second calculation group is determined by:

The technical scheme has the working principle and beneficial effects that the second calculation group is determined by calculating the most user data, the corresponding relative deviation values between the multiple physiological indexes of the volunteer user under anger emotion and the historical average values of the corresponding physiological indexes in the volunteer user normal physiological index comparison group are determined through the volunteer data investigation, the third calculation group is established according to the corresponding relative deviation values of the current multiple physiological indexes of the volunteer user, the third calculation group which is correspondingly established under anger emotion is determined, the data in the third calculation group are subjected to corresponding mean value calculation to obtain the second calculation group which is corresponding to anger emotion, the obtained average line can be determined to be the second calculation group by utilizing the data of the volunteer, the second calculation group which is corresponding to other emotion can be determined in the same manner, the third calculation group is worth noting, the second calculation group can be determined by a method which is corresponding to the mixed emotion, and the human emotion can be determined according to a preset emotion training template through the preset emotion feedback method.

In a preferred embodiment, the identifying and judging the psychological emotional state of the user through the cosine similarity algorithm by using a plurality of physiological indexes of the user further comprises:

If a certain physiological index in a certain index reference group is higher than a preset upper boundary value or lower than a preset lower boundary value corresponding to the physiological index, determining that the index reference group is abnormal, judging whether the index reference group corresponding to a plurality of subsequent moments has the same abnormality or not, and recording the occurrence times of the abnormality;

The technical scheme has the working principle and beneficial effects that another judging method is provided, namely, the psychological emotion is determined by utilizing the change condition of the physiological index, for certain emotions, the corresponding characteristics of the physiological index are changed, the change curve of the physiological index is obtained by comparing the current physiological index of a user with the corresponding preset upper boundary value and lower boundary value of the physiological index, for example, from the tense emotion to the panic emotion, the process is that the heart is accelerated to the rapid heart contraction, so that the blood pressure is instantaneously increased, the change process of the physiological index is realized, the judgment result is obtained by comparing the physiological index with the real-time data only, and therefore, the other method for judging the emotion through the linear change rule is provided, particularly, a plurality of physiological indexes obtained by analyzing the blood light reflection signal, the pulse wave signal and the electrocardio signal acquired at the same moment are divided into an index reference group, the current physiological indexes of the user are respectively compared with the corresponding preset upper boundary value and lower boundary value of the physiological index, if the physiological index existing in the certain index reference group is higher than the corresponding preset upper boundary value or lower boundary value of the physiological index, the abnormal condition is realized, the number of times of continuous abnormal conditions are prevented from being continuously recorded when the same time is high, and the number of times of abnormal conditions are continuously occur are continuously recorded, and sequentially arranging the multiple physiological indexes in the extracted multiple index reference groups after classifying according to item categories, establishing trend change curves of each physiological index, finally obtaining trend change curves corresponding to the multiple physiological indexes, such as trend change curve A corresponding to heart rate, trend change curve B corresponding to blood pressure, trend change curve C corresponding to blood oxygen saturation and the like, establishing a comparison data group (such as [ A, B, C ]) by utilizing the trend change curves corresponding to the multiple physiological indexes, taking a comparison data template corresponding to each psychological emotion, such as anger emotion, from a cloud as [ a, B, C ], respectively carrying out cosine similarity calculation on the comparison data template of each psychological emotion and the comparison data group, determining the matching degree between the comparison data group and the comparison data template corresponding to each psychological emotion, such as cosine similarity calculation on the comparison data template [ A, B, C ] and the comparison data template [ a, B, C ], and finally outputting emotion data corresponding to the greatest matching degree as emotion judgment result. The method for judging the emotion through the linear change rule is provided, and the emotion capable of enabling the physiological index to change in linear characteristics is judged.

In a preferred embodiment, the upper and lower boundary values are determined by the following method:

taking the maximum value in the second fluctuation range as a first upper boundary value, and adding a correction value to the first upper boundary value to obtain an upper boundary value;

The technical scheme has the working principle and beneficial effects that a first fluctuation range of a certain physiological index of a human body in a normal emotion state is obtained through data research of volunteers or feedback of a user group, a second fluctuation range of the certain physiological index of the human body in the normal emotion state is obtained when the current user is in the wearing equipment, a first fluctuation median value is determined according to the first fluctuation range, the first fluctuation median value is the average value of the maximum value and the minimum value of the first fluctuation range, a second fluctuation median value is determined according to the second fluctuation range, the second fluctuation median value is the average value of the maximum value and the minimum value of the second fluctuation range, the absolute difference value of the first fluctuation median value and the second fluctuation median value is determined, the absolute difference value is multiplied by a preset proportional coefficient to obtain a corrected value, the maximum value in the second fluctuation range is used as a first upper boundary value, the first upper boundary value is added with the corrected value to obtain the upper boundary value, and the minimum value in the second fluctuation range is used as a first lower boundary value, and the first lower boundary value is added with the corrected value to obtain the lower boundary value. According to the technical scheme, the deviation and trend of the first fluctuation range of a certain physiological index under the normal emotion state of the human body of the public user are considered, and the upper and lower boundary values with the characteristics of the user are established by combining the second fluctuation range of the certain physiological index under the normal emotion state acquired by the current user for many times during the use of the wearable device, so that the method is more suitable for judging the emotion condition of the user according to the physiological condition of the user.

In a preferred embodiment, the cosine similarity calculation is performed on the comparison data templates of each psychological emotion and the comparison data sets, and determining the matching degree between the comparison data sets and the comparison data templates corresponding to each psychological emotion includes:

When similarity calculation is carried out, cosine similarity calculation is carried out on all trend change curves in the comparison data set and trend change template curves of corresponding physiological indexes in the comparison data template respectively, a plurality of first-class similarities are obtained, and each first-class similarity is bound with corresponding physiological index items respectively;

and carrying out weighted calculation based on a plurality of first-class similarities according to a weight coefficient preset for each corresponding physiological index in the comparison data template of the psychological emotion, so as to obtain the matching degree between the comparison data set and the comparison data template corresponding to the psychological emotion.

The technical scheme has the working principle and beneficial effects that when cosine similarity calculation is carried out on the comparison data templates and the comparison data sets, cosine similarity calculation is needed to be carried out on all trend change curves in the comparison data sets and trend change template curves of corresponding physiological indexes in the comparison data templates respectively, so that a plurality of first-class similarities are obtained, each first-class similarity represents the similarity between the trend change curve of a certain physiological index of a user and the template trend change curve corresponding to the physiological index, after the plurality of first-class similarities are obtained, weighting calculation is carried out on the basis of the weight coefficient preset for each physiological index corresponding to the comparison data templates of the psychological emotion, so that the matching degree between the comparison data sets and the comparison data templates corresponding to the psychological emotion is obtained, and for example, the emphasis of the physiological indexes needed to be noted for judging a certain psychological emotion is different, for example, the frightened emotion is more focused on the detection on the aspect of heartbeat, and the weighted coefficient preset for the physiological indexes related to the heartbeat in the comparison data templates is more accurate, and after the psychological emotion matching calculation is carried out on the psychological emotion matching between the comparison data sets and the psychological emotion matching template.

In a preferred embodiment, the cosine similarity calculation process includes:

Determining two curves of a trend change curve and a trend change template curve;

cutting the trend change template curve with equal length according to the length of the trend change curve to obtain a second curve;

Determining the slopes of a plurality of quantized data points on a trend change curve, and arranging the slopes of the quantized data points on the trend change curve to generate a first calculation sequence;

And calculating by using the first calculation sequence and the second calculation sequence through a cosine similarity calculation formula to obtain the similarity between the trend change curve and the second curve as the first type of similarity.

The technical scheme has the working principle and beneficial effects that cosine similarity calculation is conducted on all trend change curves in a comparison data set and trend change template curves of corresponding physiological indexes in the comparison data templates respectively, the rest chord similarity calculation process comprises the steps of determining two curves of the trend change curves and the trend change template curves, conducting equal length cutting on the trend change template curves according to the lengths of the trend change curves to obtain second curves, accordingly, unequal lengths of two curves are prevented, equidistant quantization is conducted on the trend change curves and the second curves respectively to obtain a plurality of sequentially arranged quantized data points, slope of the plurality of quantized data points on the trend change curves is determined, slope arrangement of the plurality of quantized data points on the trend change curves is generated to form a first calculation sequence, slope arrangement of the plurality of quantized data points on the second curves is determined, slope arrangement of the plurality of quantized data points on the second curves is generated to form a second calculation sequence, the similarity between the trend change curves and the second curves is obtained through calculation according to the length of the trend change template curves to serve as first similarity, the slope of the quantized data points is decomposed by the curves, and the fact that the two curves are in the position of the quantized data points is calculated to represent the fact that the two curves are similar to form the quantized data points are calculated, and the fact that the two points are calculated to represent the fact that the slope of the quantized data is calculated is low in a position of the quantized data is calculated.

the signal acquisition processing module 1 is used for acquiring blood light reflection signals, pulse wave signals and electrocardiosignals of a user through wearable equipment and preprocessing the blood light reflection signals, the pulse wave signals and the electrocardiosignals to obtain human physiological signals;

the signal analysis module 2 is used for calculating and analyzing the human physiological signal value to obtain a plurality of physiological indexes of the user, wherein the physiological indexes comprise heart rate, blood oxygen saturation, atrial fibrillation and blood pressure;

The emotion judging module 3 is configured to identify and judge the psychological emotion state of the user by using a cosine similarity algorithm according to multiple physiological indexes of the user, so as to obtain an emotion judging result, and display the emotion judging result in real time on the mobile terminal.

The technical scheme has the working principle and beneficial effects that the signal acquisition processing module 1 acquires blood light reflection signals, pulse wave signals and electrocardiosignals of a user through wearable equipment such as a watch and a wrist band, performs filtering and denoising pretreatment on the blood light reflection signals, the pulse wave signals and the electrocardiosignals to obtain human physiological signals, the signal analysis module 2 performs calculation analysis on the human physiological signal values to obtain multiple physiological indexes of the user, wherein the physiological indexes comprise heart rate, blood oxygen saturation, atrial fibrillation, blood pressure and the like, and finally performs recognition judgment on psychological emotion states of the user through a cosine similarity algorithm by utilizing the multiple physiological indexes of the user through the emotion judgment module 3 to obtain emotion judgment results and display the emotion judgment results on the mobile terminal in real time. Thereby realizing continuous and uninterrupted psychological emotion insight analysis of the user.

It will be apparent to those skilled in the art that various modifications and variations can be made to the present invention without departing from the spirit or scope of the invention. Thus, it is intended that the present invention also include such modifications and alterations insofar as they come within the scope of the appended claims or the equivalents thereof.

Claims

1. A psychological and emotional insight analysis method, characterized by comprising:

The wearable device collects the user's blood light reflection signal, pulse wave signal and electrocardiogram signal, and pre-processes the blood light reflection signal, pulse wave signal and electrocardiogram signal to obtain human physiological signals;

Calculating and analyzing the human physiological signal values to obtain multiple physiological indicators of the user; wherein the physiological indicators include heart rate, blood oxygen saturation, atrial fibrillation, and blood pressure;

Using multiple physiological indicators of the user, the cosine similarity algorithm is used to identify and judge the user's psychological and emotional state, and the emotional judgment results are displayed in real time on the mobile terminal;

The method of using multiple physiological indicators of the user to identify and judge the user's psychological and emotional state through a cosine similarity algorithm includes:

Acquire multiple physiological indicators obtained by analyzing the blood light reflection signal, pulse wave signal and electrocardiogram signal collected from the user at the current moment;

Determine the historical average values of each physiological indicator of the user based on the historical physiological indicator data of the user account stored in the cloud server, and establish a normal physiological indicator control group using multiple historical average values;

Determine the relative deviation between each physiological indicator corresponding to the current moment and the historical average value of the corresponding physiological indicator in the normal physiological indicator control group;

If the relative deviation between a certain physiological indicator and the historical average value of the corresponding physiological indicator in the normal physiological indicator control group is greater than a preset deviation threshold, it is determined that the user's physiological indicator is abnormal;

When it is determined that the physiological indicators of the user are abnormal, a first calculation group is established according to a plurality of relative deviation values corresponding to the plurality of current physiological indicators of the user;

Performing cosine similarity calculations on the first calculation group and the plurality of second calculation groups stored in the cloud, respectively, to obtain similarity values between the first calculation group and the plurality of second calculation groups stored in the cloud;

Output the emotion noun bound to the second calculation group corresponding to the highest similarity value as the emotion judgment result;

While outputting the emotion judgment results, it is also necessary to calculate the intensity of the user's emotions using the following method:

Determining a plurality of relative deviation values within the second calculation group corresponding to the highest similarity value;

Calculating second relative deviation values between the plurality of relative deviation values in the first calculation group and the corresponding plurality of relative deviation values in the second calculation group;

Calculate the average of the multiple second relative deviation values to obtain the intensity of the user's emotion;

When the intensity of the user's emotions exceeds a preset fluctuation threshold, the user is reminded through the mobile terminal to pay attention to regulating his or her emotions.

2. The psychological emotion insight analysis method according to claim 1, wherein the preprocessing of the blood light reflection signal, the pulse wave signal, and the electrocardiogram signal comprises:

Digital filtering is used to remove baseline drift and ambient light noise caused by power frequency interference and respiratory jitter in blood light reflection signals, pulse waves and ECG signals;

After digital filtering, blood flow sounds in blood light reflection signals, pulse waves and ECG signals are removed through adaptive filtering;

Mean filtering and peak filtering are performed on the blood light reflection signal, pulse wave and ECG signal to eliminate the gross errors.

3. The psychological and emotional insight analysis method according to claim 1, wherein the step of calculating and analyzing the physiological signal values to obtain multiple physiological indicators of the user comprises:

Determining the user's heart rate and atrial fibrillation by calculation based on the electrocardiogram signal in the human physiological signal;

determining the user's blood oxygen saturation by calculation based on the blood light reflection signal in the human physiological signal;

The user's blood pressure is determined by calculation according to the pulse wave signal in the human physiological signal.

4. The psychological emotion insight analysis method according to claim 1, wherein the second calculation group is determined by:

Determine, through a volunteer data survey, a plurality of relative deviation values corresponding to a plurality of physiological indicators of a volunteer user under anger and a historical average value of the corresponding physiological indicators in a control group of the volunteer user with normal physiological indicators;

Establishing a third calculation group according to the multiple relative deviation values corresponding to the multiple current physiological indicators of the volunteer user;

A plurality of third calculation groups corresponding to the anger emotions of the plurality of volunteer users are determined, and corresponding mean calculations are performed on the data in the plurality of third calculation groups to obtain a second calculation group corresponding to the anger emotions.

5. The psychological and emotional insight analysis method according to claim 1, wherein the step of using multiple physiological indicators of the user to identify and judge the user's psychological and emotional state through a cosine similarity algorithm further comprises:

The multiple physiological indicators obtained by analyzing the blood light reflection signal, pulse wave signal and electrocardiogram signal collected at the same time are divided into an indicator reference group;

Pre-set upper and lower boundary values for each physiological indicator, and compare the user's current physiological indicators with their corresponding preset upper and lower boundary values;

If a physiological indicator in a certain indicator reference group is higher than the preset upper boundary value or lower than the preset lower boundary value corresponding to the physiological indicator, it is determined that the indicator reference group has an abnormality, and it is determined whether the indicator reference group corresponding to multiple consecutive moments has the same abnormality, and the number of times the abnormality occurs is recorded;

When the consecutive number of times exceeds the preset upper limit, it is determined that the user has emotional fluctuations, and multiple indicator reference groups are continuously extracted starting from the indicator reference group where the abnormality first occurs;

The multiple physiological indicators in the extracted multiple indicator reference groups are classified according to the project category and arranged in order, and a trend change curve of each physiological indicator is established, and finally the trend change curves corresponding to the multiple physiological indicators are obtained, and the trend change curves corresponding to the multiple physiological indicators are used to establish a comparative data group;

Retrieving from the cloud a comparison data template corresponding to each of the plurality of psychological emotions, performing a cosine similarity calculation on the comparison data template for each psychological emotion and the comparison data group, and determining a degree of match between the comparison data group and the comparison data template corresponding to each psychological emotion;

The psychological emotion corresponding to the comparison data template with the greatest matching degree is output as the emotion judgment result.

6. The psychological emotion insight analysis method according to claim 5, wherein the upper boundary value and the lower boundary value are determined by the following method:

Obtain the first fluctuation range of a physiological indicator of the human body under normal emotional state through self-volunteer data survey or user group feedback;

Obtain the second fluctuation range of a physiological indicator of the current user in a normal emotional state while using the wearable device;

Determine a first fluctuation median value according to the first fluctuation range, where the first fluctuation median value is the average of the maximum value and the minimum value of the first fluctuation range;

determining a second fluctuation median according to the second fluctuation range, where the second fluctuation median is the average of the maximum value and the minimum value of the second fluctuation range;

determining an absolute difference between the first fluctuation median and the second fluctuation median, and multiplying the absolute difference by a preset proportional coefficient to obtain a correction value;

The minimum value in the second fluctuation range is used as the first lower boundary value, and the correction value is added to the first lower boundary value to obtain the lower boundary value.

7. The psychological emotion insight analysis method according to claim 5, wherein the step of performing cosine similarity calculations on the comparison data templates for each psychological emotion and the comparison data group to determine the degree of match between the comparison data group and the comparison data templates corresponding to each psychological emotion comprises:

Select a comparison data template of a certain psychological emotion and perform similarity calculation between it and the comparison data group;

When performing similarity calculation, cosine similarity calculation is performed on all trend change curves in the comparison data group and trend change template curves of corresponding physiological indicators in the comparison data template to obtain multiple first-class similarities, and each first-class similarity is respectively bound to its corresponding physiological indicator item; wherein, the cosine similarity calculation process includes:

Determine two curves, the trend change curve and the trend change template curve;

According to the length of the trend change curve, the trend change template curve is trimmed to equal length to obtain a second curve;

quantizing the trend change curve and the second curve at equal distances to obtain a plurality of sequentially arranged quantified data points;

determining slopes of a plurality of quantified data points on the trend change curve, and arranging the slopes of the plurality of quantified data points on the trend change curve to generate a first calculation sequence;

determining slopes of a plurality of quantized data points on the second curve, and arranging the slopes of the plurality of quantized data points on the second curve to generate a second calculation sequence;

Using the first calculation sequence and the second calculation sequence, calculating the similarity between the trend change curve and the second curve using a cosine similarity calculation formula as a first type of similarity;

According to the weight coefficient preset for each physiological indicator corresponding to the comparison data template of the psychological emotion, a weighted calculation is performed based on multiple first-category similarities to obtain the degree of matching between the comparison data group and the comparison data template corresponding to the psychological emotion.

8. A psychological and emotional insight analysis system, comprising:

The signal acquisition and processing module is used to collect the user's blood light reflection signal, pulse wave signal and electrocardiogram signal through the wearable device, and pre-process the blood light reflection signal, pulse wave signal and electrocardiogram signal to obtain human physiological signals;

A signal analysis module, configured to calculate and analyze the human physiological signal values to obtain a plurality of physiological indicators of the user; wherein the physiological indicators include heart rate, blood oxygen saturation, atrial fibrillation, and blood pressure;

The emotion judgment module is used to use multiple physiological indicators of the user to identify and judge the user's psychological and emotional state through the cosine similarity algorithm to obtain the emotion judgment result and display it in real time on the mobile terminal;

The method in which the emotion judgment module uses multiple physiological indicators of the user to identify and judge the user's psychological and emotional state through a cosine similarity algorithm includes: