CN115221646A - Purifier Turbulence Fitting System, Evaluation System and Control System - Google Patents

Abstract

The invention relates to a purifier turbulence fitting system, an intelligent state evaluation system and an intelligent control system based on an antagonistic neural network, wherein the intelligent control system comprises a cloud purifier turbulence vector pre-training model, a purifier multipoint bit flow rate sensor, a neural network turbulence vector correction module, a turbulence vector pre-processing model, a single time slice generation module for generating a neural network turbulence vector, a single time slice evaluation module for judging the neural network turbulence vector, a multi-time slice generation module for generating the neural network turbulence vector, a multi-time slice evaluation module for judging the neural network turbulence vector, a neural network turbulence vector fitting flow direction module, an operation state evaluation module for the neural network purifier and the intelligent control system. The invention uses the common detection of a plurality of detection point positions and a pre-training model of the purifier to fit and iteratively correct the turbulent output of the purifier, thereby realizing the accurate control of the purifier.

Description

Purifier Turbulence Fitting System, Evaluation System and Control System

technical field

The invention relates to the field of purifier intelligent control, in particular to a purifier turbulence fitting system, an intelligent state evaluation system and an intelligent control system based on an adversarial neural network.

Background technique

In the public space air purification scheme, the traditional method is to read the fixed point value. When there is an excessive value, the purifier is controlled to purify at a fixed angle and a fixed air volume. More advanced ones also use intelligent control to automatically adjust the parameters of the purifier. However, the previous intelligent adjustment method did not fit the key output turbulence model, and thus could not effectively fit the actual purification efficiency of the purifier, could not effectively control the purifier intelligently, and could not judge the working quality of the current equipment through turbulence fitting. Thereby prompting maintenance and replacement.

SUMMARY OF THE INVENTION

In view of the problems existing in the prior art, the purpose of the present invention is to provide a purifier turbulence fitting system, an intelligent state evaluation system and an intelligent control system based on an adversarial neural network, so as to realize more effective and accurate control of the purifier.

For achieving the above object, the technical scheme adopted in the present invention is:

A purifier turbulence fitting system based on adversarial neural network, which includes

The purifier turbulence vector pre-training model is used to simulate and output the turbulence vector model according to the operating conditions of the purifier before leaving the factory;

The neighborhood air quality monitor cluster consists of multiple air quality monitors installed at different locations of the purifier, which are used to collect air pollutant concentrations, and each air pollutant concentration is accompanied by location information;

The multi-point flow velocity sensor of the purifier is composed of a plurality of flow velocity sensors arranged at different positions of the purifier, which are used to collect the real-time air flow velocity of the position points;

The neural network turbulence vector correction module is connected to the cloud purifier turbulence vector pre-training model, the domain air quality monitor cluster, and the purifier multi-point flow velocity sensor, which is used to input the turbulence vector, air pollutant concentration with location information, and multiple The real-time air velocity at the location point; the neural network turbulence vector correction module corrects the input turbulent vector model according to the input air pollutant concentration and real-time air velocity, and obtains the turbulent vector preprocessing model;

The generating neural network turbulence vector single time slice generation module is connected to the neighborhood air quality monitor cluster and the purifier multi-point flow velocity sensor. The generating neural network turbulence vector single time slice generation module is used to input the air pollutant concentration with location information. , real-time air velocity at multiple locations, and generate multiple single-time slice turbulence vector pre-fitting vectors according to the input information;

Discriminant neural network turbulence vector single time slice evaluation module, connected to turbulence vector preprocessing model, generating neural network turbulence vector single time slice generation module, used to input turbulence vector preprocessing model and multiple single time slice turbulence vector pre-fitting vectors , and score each turbulence vector pre-fitting vector through the neural network, select the highest scoring turbulence vector as the single time slice turbulence vector and output it;

Generate neural network turbulence vector multi-time slice generation module, connect the discriminative neural network turbulence vector single time slice evaluation module, input the turbulence vector of the current single time slice, and generate the neural network according to the current single time slice turbulence vector and multiple previous time slices. Turbulence vector, generating multiple multi-time slice turbulence vector pre-fitting vectors;

The discriminant neural network turbulence vector multi-time slice evaluation module is used to input the turbulence vector preprocessing model, multiple multi-time slice turbulence vector pre-fitting vectors, and multiple time slice turbulence vector pre-fitting vectors corresponding to multiple single time slices. Turbulence vector score, score each multi-time slice turbulence vector pre-fitting vector through the trained neural network, and select the highest score as the multi-time slice corrected turbulence vector of a single time slice and output it;

The neural network turbulence vector fitting flow direction module is connected to the discriminative neural network turbulence vector multi-time slice evaluation module, which is used to input the multi-time slice corrected turbulence vector.

The fitting system also includes

Generating Neural Network Turbulence Vector Single Time Slice Generation Module The real-time training module is connected to the Generating Neural Network Turbulence Vector Single Time Slice Generation Module and the Discriminant Neural Network Turbulence Vector Single Time Slice Evaluation Module, and the input generates a neural network turbulence vector single time slice at intervals. The current network weight of the slice generation module, and the evaluation score of the turbulence vector fitting vector of the corresponding time slice given by all the discriminative neural network turbulence vector single-time slice evaluation modules in this time interval, training to generate neural network turbulence vector single-time The network weights of the slice generation module and overwrite the original network weights.

The fitting system also includes

Generating neural network turbulence vector multi-time slice generation module The real-time training module is connected to the discriminative neural network turbulence vector multi-time slice evaluation module, and the current network weight of the neural network turbulence vector multi-time slice generation module is input at intervals, which is the same as the current time. The evaluation scores of the turbulence vector fitting vectors corresponding to multiple time slices given by the multi-time slice evaluation module of all discriminative neural network turbulence vectors in the interval are trained to generate the network weights of the neural network turbulence vector multi-time slice generation module, and cover the original network weight.

A purifier intelligent state assessment system based on adversarial neural network, which includes

The purifier turbulence vector pre-training model is used to simulate and output the turbulence vector model according to the operating conditions of the purifier before leaving the factory;

The neighborhood air quality monitor cluster consists of multiple air quality monitors installed at different locations of the purifier, which are used to collect air pollutant concentrations, and each air pollutant concentration is accompanied by location information;

The multi-point flow velocity sensor of the purifier is composed of a plurality of flow velocity sensors arranged at different positions of the purifier, which are used to collect the real-time air flow velocity of the position points;

The neural network turbulence vector correction module is connected to the cloud purifier turbulence vector pre-training model, the domain air quality monitor cluster, and the purifier multi-point flow velocity sensor, which is used to input the turbulence vector, air pollutant concentration with location information, and multiple The real-time air velocity at the location point; the neural network turbulence vector correction module corrects the input turbulent vector model according to the input air pollutant concentration and real-time air velocity, and obtains the turbulent vector preprocessing model;

The generating neural network turbulence vector single time slice generation module is connected to the neighborhood air quality monitor cluster and the purifier multi-point flow velocity sensor. The generating neural network turbulence vector single time slice generation module is used to input the air pollutant concentration with location information. , real-time air velocity at multiple locations, and generate multiple single-time slice turbulence vector pre-fitting vectors according to the input information;

Discriminant neural network turbulence vector single time slice evaluation module, connected to turbulence vector preprocessing model, generating neural network turbulence vector single time slice generation module, used to input turbulence vector preprocessing model and multiple single time slice turbulence vector pre-fitting vectors , and score each turbulence vector pre-fitting vector through the neural network, select the highest scoring turbulence vector as the single time slice turbulence vector and output it;

Generate neural network turbulence vector multi-time slice generation module, connect the discriminative neural network turbulence vector single time slice evaluation module, input the turbulence vector of the current single time slice, and generate the neural network according to the current single time slice turbulence vector and multiple previous time slices. Turbulence vector, generating multiple multi-time slice turbulence vector pre-fitting vectors;

The discriminant neural network turbulence vector multi-time slice evaluation module is used to input the turbulence vector preprocessing model, multiple multi-time slice turbulence vector pre-fitting vectors, and multiple time slice turbulence vector pre-fitting vectors corresponding to multiple single time slices. Turbulence vector score, score each multi-time slice turbulence vector pre-fitting vector through the trained neural network, and select the highest score as the multi-time slice corrected turbulence vector of a single time slice and output it;

The neural network purifier operation status evaluation module is connected to the discriminative neural network turbulence vector multi-time slice evaluation module, which is used to input the multi-time slice correction turbulence vector.

The state assessment system also includes

Generating Neural Network Turbulence Vector Single Time Slice Generation Module The real-time training module is connected to the Generating Neural Network Turbulence Vector Single Time Slice Generation Module and the Discriminant Neural Network Turbulence Vector Single Time Slice Evaluation Module, and the input generates a neural network turbulence vector single time slice at intervals. The current network weight of the slice generation module, and the evaluation score of the turbulence vector fitting vector of the corresponding time slice given by all the discriminative neural network turbulence vector single-time slice evaluation modules in this time interval, training to generate neural network turbulence vector single-time The network weights of the slice generation module and overwrite the original network weights.

The state assessment system also includes

Generating neural network turbulence vector multi-time slice generation module The real-time training module is connected to the discriminative neural network turbulence vector multi-time slice evaluation module, and the current network weight of the neural network turbulence vector multi-time slice generation module is input at intervals, which is the same as the current time. The evaluation scores of the turbulence vector fitting vectors corresponding to multiple time slices given by the multi-time slice evaluation module of all discriminative neural network turbulence vectors in the interval are trained to generate the network weights of the neural network turbulence vector multi-time slice generation module, and cover the original network weight.

An intelligent control system for a purifier based on an adversarial neural network, which includes the above-mentioned turbulence fitting system and the above-mentioned intelligent state assessment system;

The intelligent control system also includes an intelligent control module, which is connected to the neural network purifier operating condition evaluation module and the neural network turbulence vector flow direction module, and is used to input the purifier operating state evaluation value and the neighborhood turbulence vector diagram, and operate according to the purifier. The state evaluation value and the neighborhood turbulence vector diagram adjust the working mode of the purifier.

After adopting the above scheme, the present invention uses the joint detection of multiple detection points, and the pre-training model of the purifier to fit and iteratively correct the turbulent output of the purifier, thereby realizing accurate control of the purifier. The invention not only considers the turbulence vector of a single time slice, but also considers other single time slices before and after the single time slice on this basis, so that the turbulence vector of multiple time slices makes the formed turbulence vector more accurate, and then can accurately control the purifier . Specifically, the present invention uses the concept of single time slice to fit the turbulence vector considered by the neural network, and then uses the concept of multiple time slices to pick out the situation that the turbulence vector changes with time from these generated vectors. To put it simply, the turbulence vector is actually generated in a single time slice, and a selection process is performed for multiple time slices, and then the selection result is compared with the final sensor data to obtain an accurate turbulence vector.

Description of drawings

FIG. 1 is a principle block diagram of the present invention.

Detailed ways

As shown in Figure 1, the present invention discloses an intelligent control system for a purifier based on an adversarial neural network, which includes:

The purifier turbulence vector pre-training model is used to simulate and output the turbulence vector model according to the operating conditions of the purifier before leaving the factory. The purifier turbulence vector pre-training model is currently set up in the cloud, and can also be set up locally.

The neighborhood air quality monitor cluster consists of a plurality of air quality monitors arranged at different positions of the purifier, which are used to collect air pollutant concentrations, and each air pollutant concentration is accompanied by location information.

The multi-point flow velocity sensor of the purifier is composed of a plurality of flow velocity sensors arranged at different positions of the purifier, which are used to collect the real-time air flow velocity of the position points.

The neural network turbulence vector correction module is connected to the cloud purifier turbulence vector pre-training model, the domain air quality monitor cluster, and the purifier multi-point flow velocity sensor, which is used to input the turbulence vector, air pollutant concentration with location information, and multiple The real-time air velocity at the location point; the neural network turbulence vector correction module corrects the input turbulence vector model according to the input air pollutant concentration and real-time air velocity, and obtains the turbulence vector preprocessing model.

A neural network turbulence vector single time slice generation module is connected to the neighborhood air quality monitor cluster and the purifier multi-point flow velocity sensor, which is used to input the air pollutant concentration with location information and the real-time air flow velocity of multiple location points. According to the input information, the network generates multiple single-time slice turbulence vector pre-fitting vectors.

Discriminant neural network turbulence vector single time slice evaluation module, connected to turbulence vector preprocessing model, generating neural network turbulence vector single time slice generation module, used to input turbulence vector preprocessing model and multiple single time slice turbulence vector pre-fitting vectors , and score each turbulence vector pre-fitting vector through the neural network, and select the highest scoring turbulence vector as the single time slice turbulence vector and output it.

Generating Neural Network Turbulence Vector Single Time Slice Generation Module The real-time training module is connected to the Generating Neural Network Turbulence Vector Single Time Slice Generation Module and the Discriminant Neural Network Turbulence Vector Single Time Slice Evaluation Module, and the input generates a neural network turbulence vector single time slice at intervals. The current network weight of the slice generation module, and the evaluation score of the turbulence vector fitting vector of the corresponding time slice given by all the discriminative neural network turbulence vector single-time slice evaluation modules in this time interval, training to generate neural network turbulence vector single-time The network weights of the slice generation module and overwrite the original network weights.

Generate neural network turbulence vector multi-time slice generation module, connect the discriminative neural network turbulence vector single time slice evaluation module, input the turbulence vector of the current time slice, and generate the neural network according to the turbulence vector of the current time slice and the turbulence vector of multiple previous time slices , to generate multiple multi-time slice turbulence vector pre-fit vectors.

The discriminant neural network turbulence vector multi-time slice evaluation module is characterized by the input cloud turbulence vector preprocessing model, multiple multi-time slice turbulence vector pre-fitting vectors, and multiple single time slices corresponding to the multi-time slice turbulence vector pre-fitting vectors Each multi-time slice turbulence vector pre-fitting vector is scored through the trained neural network, and the highest score is selected as the multi-time slice corrected turbulence vector of a single time slice and output.

Generating neural network turbulence vector multi-time slice generation module The real-time training module is connected to the discriminative neural network turbulence vector multi-time slice evaluation module, and the current network weight of the neural network turbulence vector multi-time slice generation module is input at intervals, which is the same as the current time. The evaluation scores of the turbulence vector fitting vectors corresponding to multiple time slices given by the multi-time slice evaluation module of all discriminative neural network turbulence vectors in the interval are trained to generate the network weights of the neural network turbulence vector multi-time slice generation module, and cover the original network weight.

The neural network turbulence vector fitting flow direction module is connected to the discriminative neural network turbulence vector multi-time slice evaluation module, which is used to input the multi-time slice corrected turbulence vector.

The neural network purifier operation status evaluation module is connected to the discriminative neural network turbulence vector multi-time slice evaluation module, which is used to input the multi-time slice correction turbulence vector. The neural network purifier operating condition evaluation module compares the multi-time slice corrected turbulence vector with the historical multi-time vector to see if the specific output has changed significantly under similar conditions, because generally large changes indicate problems .

The intelligent control module is connected to the neural network purifier operating condition evaluation module and the neural network turbulence vector flow direction module, which is used to input the purifier operating state evaluation value and the neighborhood turbulence vector diagram, and according to the purifier operating state evaluation value and the neighborhood turbulence vector Figure to adjust the working mode of the purifier.

The above cloud purifier turbulence vector pre-training model, purifier multi-point flow velocity sensor, neural network turbulence vector correction module, turbulence vector preprocessing model, generating neural network turbulence vector single time slice generation module, discriminant neural network turbulence vector single time slice Evaluation Module, Generative Neural Network Turbulence Vector Single Time Slice Generation Module Real-time Training Module, Generative Neural Network Turbulence Vector Multi-Time Slice Generation Module, Discriminant Neural Network Turbulence Vector Multi-Time Slice Evaluation Module, Generative Neural Network Turbulence Vector Multi-Time Slice Generation Module Real-time The training module and the neural network turbulence vector fitting flow direction module can constitute a purifier turbulence fitting system based on an adversarial neural network, which is used to obtain the neighborhood turbulence vector diagram of the purifier.

The above cloud purifier turbulence vector pre-training model, purifier multi-point flow velocity sensor, neural network turbulence vector correction module, turbulence vector preprocessing model, generating neural network turbulence vector single time slice generation module, discriminant neural network turbulence vector single time slice Evaluation Module, Generative Neural Network Turbulence Vector Single Time Slice Generation Module Real-time Training Module, Generative Neural Network Turbulence Vector Multi-Time Slice Generation Module, Discriminant Neural Network Turbulence Vector Multi-Time Slice Evaluation Module, Generative Neural Network Turbulence Vector Multi-Time Slice Generation Module Real-time The training module and the neural network purifier operating condition evaluation module can constitute a purifier intelligent state evaluation system based on the confrontation neural network, which is used to obtain the purifier operating state evaluation value.

The above are only the embodiments of the present invention and do not limit the technical scope of the present invention. Therefore, any minor modifications, equivalent changes and modifications made to the above embodiments according to the technical essence of the present invention still belong to the present invention. within the scope of the technical solution.

Claims (7)

1. Purifier torrent fitting system based on antagonizing neural network, its characterized in that: the system comprises

The purifier turbulence vector pre-training model is used for simulating and outputting a turbulence vector model according to the running condition of the purifier before leaving the factory;

the neighborhood air quality monitor cluster consists of a plurality of air quality monitors arranged at different position points of the purifier and is used for collecting the concentration of air pollutants, and each air pollutant concentration is accompanied by position information;

the multi-point flow velocity sensor of the purifier consists of a plurality of flow velocity sensors which are arranged at different position points of the purifier and are used for collecting the real-time air velocity of the position points;

the neural network turbulence vector correction module is connected with the cloud purifier turbulence vector pre-training model, the field air quality monitor cluster and the purifier multipoint bit flow rate sensor and is used for inputting turbulence vectors, air pollutant concentrations with position information and real-time air flow rates of a plurality of position points; the neural network turbulence vector correction module corrects the input turbulence vector model according to the input air pollutant concentration and the real-time air flow rate to obtain a turbulence vector pretreatment model;

the generating neural network turbulence vector single time slice generating module is connected with the neighborhood air quality monitor cluster and the purifier multipoint bit flow rate sensor, and is used for inputting the air pollutant concentration with position information and the real-time air flow rate of a plurality of position points and generating a plurality of single time slice turbulence vector pre-fitting vectors according to the input information;

the system comprises a turbulence vector single-time-slice evaluation module for judging a neural network, a turbulence vector preprocessing model, a generation neural network turbulence vector single-time-slice generation module, a judgment module and a prediction module, wherein the turbulence vector single-time-slice evaluation module is connected with the turbulence vector preprocessing model and the generation neural network turbulence vector single-time-slice generation module and is used for inputting the turbulence vector preprocessing model and a plurality of single-time-slice turbulence vector pre-fitting vectors, scoring each turbulence vector pre-fitting vector through the neural network, and selecting the turbulence vector with the highest score as a single time slice from the turbulence vectors and outputting the turbulence vector;

generating a neural network turbulence vector multi-time slice generating module, connecting and judging the neural network turbulence vector single-time slice evaluating module, inputting the turbulence vector of the current single time slice, generating a plurality of multi-time slice turbulence vector pre-fitting vectors by the neural network according to the turbulence vector of the current single time slice and the turbulence vectors of a plurality of previous time slices;

the multi-time-slice evaluation module is used for inputting a turbulence vector preprocessing model, a plurality of multi-time-slice turbulence vector pre-fitting vectors and turbulence vector scores of a plurality of single time slices corresponding to the multi-time-slice turbulence vector pre-fitting vectors, scoring each multi-time-slice turbulence vector pre-fitting vector through the trained neural network, and selecting and outputting a multi-time-slice correction turbulence vector with the highest score as the single time slice;

the neural network turbulence vector fitting flow direction module is connected with the neural network turbulence vector multi-time-slice evaluation module and used for inputting multi-time-slice correction turbulence vectors, and the neural network corrects the turbulence vectors according to the multi-time-slices and outputs a neighborhood turbulence vector diagram.

2. The anti-neural-network based purifier turbulence fitting system of claim 1, wherein: the fitting system further comprises

The real-time training module is connected with the generating module for generating the single time slice of the neural network turbulence vector and the evaluating module for judging the single time slice of the neural network turbulence vector, the current network weight of the generating module for generating the single time slice of the neural network turbulence vector is input at intervals, and the evaluation scores of the turbulence vector fitting vectors of the corresponding time slices given by all the evaluating modules for judging the single time slice of the neural network turbulence vector in the time interval are trained to generate the network weight of the generating module for generating the single time slice of the neural network turbulence vector, and the original network weight is covered.

3. The anti-neural-network-based purifier turbulence-fitting system of claim 1, wherein: the fitting system further comprises

The real-time training module is connected with the judgment neural network turbulence vector multi-time-slice evaluation module, the current network weight of the generation neural network turbulence vector multi-time-slice generation module is input at intervals, and the evaluation scores of turbulence vector fitting vectors corresponding to multiple time slices given by all judgment neural network turbulence vector multi-time-slice evaluation modules in the time interval are trained to generate the network weight of the neural network turbulence vector multi-time-slice generation module and cover the original network weight.

4. Purifier intelligence state evaluation system based on antagonizing neural network, its characterized in that: the system comprises

The purifier turbulence vector pre-training model is used for simulating and outputting a turbulence vector model according to the running condition of the purifier before leaving the factory;

the neighborhood air quality monitor cluster consists of a plurality of air quality monitors arranged at different position points of the purifier and is used for collecting the concentration of air pollutants, and each air pollutant concentration is accompanied by position information;

the purifier multipoint flow velocity sensor consists of a plurality of flow velocity sensors which are arranged at different position points of the purifier and are used for collecting the real-time air velocity of the position points;

the neural network turbulence vector correction module is connected with the cloud purifier turbulence vector pre-training model, the field air quality monitor cluster and the purifier multi-point bit flow rate sensor and is used for inputting turbulence vectors, air pollutant concentrations with position information and real-time air flow rates of a plurality of position points; the neural network turbulence vector correction module corrects the input turbulence vector model according to the input air pollutant concentration and the real-time air flow rate to obtain a turbulence vector pretreatment model;

the generating neural network turbulence vector single time slice generating module is connected with the neighborhood air quality monitor cluster and the purifier multipoint bit flow rate sensor, and is used for inputting the air pollutant concentration with position information and the real-time air flow rate of a plurality of position points and generating a plurality of single time slice turbulence vector pre-fitting vectors according to the input information;

the system comprises a turbulence vector single-time-slice evaluation module for judging a neural network, a turbulence vector preprocessing model, a generation neural network turbulence vector single-time-slice generation module, a judgment module and a prediction module, wherein the turbulence vector single-time-slice evaluation module is connected with the turbulence vector preprocessing model and the generation neural network turbulence vector single-time-slice generation module and is used for inputting the turbulence vector preprocessing model and a plurality of single-time-slice turbulence vector pre-fitting vectors, scoring each turbulence vector pre-fitting vector through the neural network, and selecting the turbulence vector with the highest score as a single time slice from the turbulence vectors and outputting the turbulence vector;

generating a neural network turbulence vector multi-time slice generating module, connecting and judging the neural network turbulence vector single-time slice evaluating module, inputting the turbulence vector of the current single time slice, generating a plurality of multi-time slice turbulence vector pre-fitting vectors by the neural network according to the turbulence vector of the current single time slice and the turbulence vectors of a plurality of previous time slices;

the multi-time-slice evaluation module is used for inputting a turbulence vector preprocessing model, a plurality of multi-time-slice turbulence vector pre-fitting vectors and turbulence vector scores of a plurality of single time slices corresponding to the multi-time-slice turbulence vector pre-fitting vectors, scoring each multi-time-slice turbulence vector pre-fitting vector through the trained neural network, and selecting and outputting a multi-time-slice correction turbulence vector with the highest score as the single time slice;

the neural network purifier operation condition evaluation module is connected with the neural network turbulence vector multi-time slice evaluation module and used for inputting multi-time slice correction turbulence vectors, and the neural network corrects the turbulence vectors according to the multi-time slices and outputs purifier operation condition evaluation values.

5. The anti-neural-network-based intelligent state evaluation system for purifiers according to claim 4, wherein: the state evaluation system further comprises

The real-time training module is connected with the generation module for generating the single time slice of the neural network turbulence vector and the evaluation module for judging the single time slice of the neural network turbulence vector, inputs the current network weight of the generation module for generating the single time slice of the neural network turbulence vector at intervals, and trains the network weight of the generation module for generating the single time slice of the neural network turbulence vector according to the evaluation scores of the turbulence vector fitting vectors of the corresponding time slices given by the evaluation modules for judging the single time slice of the neural network turbulence vector in the current time interval, and covers the original network weight.

6. The anti-neural-network-based intelligent state evaluation system for purifiers according to claim 4, wherein: the state evaluation system further comprises

The real-time training module is connected with the judgment neural network turbulence vector multi-time-slice evaluation module, the current network weight of the generation neural network turbulence vector multi-time-slice generation module is input at intervals, and the evaluation scores of turbulence vector fitting vectors corresponding to multiple time slices given by all judgment neural network turbulence vector multi-time-slice evaluation modules in the time interval are trained to generate the network weight of the neural network turbulence vector multi-time-slice generation module and cover the original network weight.

7. Purifier intelligence control system based on antagonism neural network, its characterized in that: the system comprises a turbulence fitting system as claimed in any one of claims 1 to 3 and an intelligent state assessment system as claimed in any one of claims 4 to 6;

the intelligent control system also comprises an intelligent control module which is connected with the neural network purifier running state evaluation module and the neural network turbulence vector flow direction module and is used for inputting the purifier running state evaluation value and the neighborhood turbulence vector diagram and adjusting the working mode of the purifier according to the purifier running state evaluation value and the neighborhood turbulence vector diagram.

Images