CN104809338A - Satellite in orbit space-environment-influence early warning method based on correlation relationship - Google Patents

Abstract

The present invention provides an early warning method based on the relationship between satellites in orbit affected by the space environment, which includes: obtaining the time-domain correlation Xo(T) between the two whose reliability and correlation exceed the set threshold; obtaining satellite The quantitative index G(t) of the computer SRAM device in the star mission center affected by the space radiation environment at time t; multiple supportability values and multiple credibility values are obtained according to the correlation between G(t) and A(t), Select an association relationship with the largest support value among multiple support values. If the support value is greater than 0.6, the alarm information is "yes" and the confidence value corresponding to the support value is the warning value Y( t), if the support value is less than or equal to 0.6, the alarm information is "none". The present invention can carry out early warning of satellite on-orbit anomalies through the analysis of the relationship between single event effects and satellite on-orbit anomalies and according to the real-time data of space-borne high-energy particle flux detection equipment, and has the advantages of high reliability, pertinence and accuracy.

Description

An Early Warning Method Based on Correlation Relationships for In-orbit Satellites Affected by Space Environment

technical field

The invention belongs to the technical field of on-orbit operation management of spacecraft, and in particular relates to an early warning method based on an association relationship for an on-orbit satellite to be affected by a space environment.

Background technique

Statistics at home and abroad show that the space radiation environment caused by the Earth's radiation belt (hereinafter referred to as the radiation belt) is one of the main causes of abnormal spacecraft failures. Among the more than 5,000 fault records collected by the US NGDC (National Geophysical Data Center) during 1966-1994, about half were diagnosed as caused by the space environment, and about 2,300 of them were caused by the space radiation environment. The statistics of Hecht and Remez et al. in the United States in 1996 showed that "among the failures and abnormalities of spacecraft in the past 40 years, more than 40% were caused by design and space environment in the first 30 years, and 36% were still in the last 10 years." Currently. , my country's spacecraft are affected by the space radiation environment and cause failures from time to time.

Analysis and research show that the root cause of spacecraft failure is various space radiation effects caused by the space radiation environment. It mainly includes: the single event effect that a single high-energy particle causes device logic inversion or latch-up, the charging and discharging effect caused by the accumulation of high-energy electrons inside the spacecraft, the radiation dose effect caused by high-energy particles through ionization and atomic displacement, etc. These effects can directly cause on-board computer false instructions, material performance degradation of spacecraft or component parts, device lock-up and burnout, lead to control computer disorder, spacecraft attitude loss of control or reversal, lack of energy, and even lead to life reduction or failure of the spacecraft. It has greatly affected the effectiveness of the aerospace equipment system.

Therefore, to effectively evaluate the impact of space radiation environmental effects on spacecraft on-orbit operation, to study the correlation between space radiation environmental effects and spacecraft on-orbit anomalies, and to establish an early warning method for on-orbit spacecraft based on the mechanism of space radiation environmental effects can be based on Space environment forecast results, systematically assess the hazards of on-orbit spacecraft affected by the space environment, and give early warnings of possible failures, abnormalities or state transitions of on-orbit spacecraft after space environment events occur, through on-orbit Safe operations are handled in a timely manner, or measures are taken to prevent the recurrence of abnormalities, so as to ensure the safe and reliable operation of spacecraft in orbit and the successful completion of space missions.

At present, in the field of engineering application and research, the fault diagnosis and early warning technology and method research work is mainly carried out based on the telemetry parameters transmitted by the spacecraft, and the research institutions related to the space environment also make predictions based on the space environment information.

1 Fault diagnosis and early warning method based on spacecraft telemetry parameters

1) Fault diagnosis method based on signal processing. This method is one of the earlier methods in the field of diagnosis, mainly using the threshold model. Signal analysis mainly uses time domain, frequency domain, amplitude, and time-frequency domain characteristic analysis. Signal processing methods mainly include: peak value, root mean square value, crest coefficient, form coefficient, skewness index and other parameter analysis, correlation analysis method, envelope analysis method, maximum entropy spectrum method, cepstrum method, synchronous signal averaging method, Autoregressive spectral analysis, wavelet analysis, fractal analysis, etc. Signal analysis methods are the basis for other diagnostic methods.

2) A rule-based expert system diagnosis method. The rule-based method is also called the production method. The early fault diagnosis expert systems are all based on rules. These rules are summarized from the experience of experts and used to describe the relationship between faults and symptoms. The advantage of this method is that the knowledge representation is simple, intuitive, visual, and convenient. It uses direct knowledge representation and relatively simple heuristic knowledge, and the diagnosis and reasoning speed is fast; the storage space of the required data is relatively small; it is easy to program and develop a rapid prototype. system. The disadvantage is that the fault modes covered by the knowledge base are limited, and it is powerless to diagnose the faults that have not occurred before and are inexperienced; when there is no corresponding rule matching the symptoms in the knowledge base, it is easy to cause misdiagnosis or diagnosis failure.

3) Diagnosis method based on fault tree. The fault tree is a directed graph that reflects the fault propagation relationship. It takes the most undesirable event of the diagnostic object as the top event, and expands layer by layer according to the structure and functional relationship of the object until the inseparable event (bottom event). Its advantage is that it can realize rapid diagnosis; the knowledge base can be easily modified dynamically and can maintain consistency; probabilistic reasoning can be used to select the search channel of rules to a certain extent to improve the efficiency of diagnosis; Given the fault tree of , the diagnosis can be realized. The disadvantage is that because the fault tree is based on component connection and failure mode analysis, unpredictable faults cannot be diagnosed; the diagnosis result depends heavily on the completeness of the fault tree information.

4) Other methods, such as neural network-based diagnosis methods, model-based fault diagnosis methods, Petri net-based fault diagnosis methods, multi-sensor information fusion fault diagnosis methods, etc.

The existing fault early warning methods that can be checked are mainly to predict the future trend of spacecraft telemetry parameters, and substitute the predicted parameter values into the fault diagnosis system for diagnosis, and the obtained diagnosis results are the early warning results.

2 Early warning method based on space environment information

Domestic institutions specializing in space environment research (such as the Space Center of the Chinese Academy of Sciences, Purple Mountain Observatory, etc.) combine space environment models and data, based on observed solar activities, to predict and forecast space environment events and geomagnetic activity levels in the future. According to different levels of proton events, electronic events, and geomagnetic storms and other events, a universal early warning is provided for the safety of spacecraft operating in orbit.

The early warning method based on space environment information also has certain limitations. First of all, most of the space environment information obtained by space environment monitoring agencies currently describes the average value on a global scale, and there are large differences in the space environment experienced by spacecraft in different orbits. , it is difficult to obtain the precise value of the orbit of a certain spacecraft only by means of the averaged space environment data; secondly, due to differences in the degree of radiation resistance design of the spacecraft, the single event turnover threshold and the total dose of radiation resistance of the selected components are different, and the space The mechanism and degree of environmental impact on spacecraft are also different. It is difficult to quantitatively evaluate the degree of impact on a spacecraft from the perspective of space environment alone. Therefore, early warning methods based on space environment data have limited effects.

The above methods all start from the change law of the spacecraft telemetry data, and study the diagnosis and early warning methods. However, in the actual on-orbit management work, it is found that 80% of the anomalies are triggered by space environmental factors, and the telemetry parameters associated with many anomalies will mutate when anomalies occur, which is difficult to discover through telemetry parameter prediction. Therefore, for such anomalies, the existing early warning methods are less effective.

Contents of the invention

In order to solve the above problems, the present invention provides a correlation-based early warning method for on-orbit satellites affected by the space environment, which analyzes the correlation between single event effects and satellite in-orbit anomalies and based on real-time data from spaceborne high-energy particle flux detection equipment. Early warning of satellite in-orbit abnormalities has the advantages of reliability, pertinence and high accuracy.

The early warning method that the on-orbit satellite is affected by the space environment based on the association relationship of the present invention comprises:

Step 1. Analyze the relationship between particle effects and satellite in-orbit anomalies, and obtain the time-domain relationship Xo(T) between the two when the reliability and correlation exceed the set threshold:

Step 11: Carry out pre-processing of the detection equipment data of the spaceborne high-energy particle flux detection equipment carried by satellites in a certain historical period T, and then convert it into the high-energy particle flux of the space where T satellites are located in a certain historical period T Lo(T);

Among them, the wild preprocessing includes three steps:

The first step: Eliminate outliers that exceed the data range of the detection equipment;

Step 2: Perform data screening according to the type of detection equipment data and the location of the sub-satellite point corresponding to the data collection time:

When the data of the detection equipment is proton data: For the proton data at the time of data collection and the position of the sub-satellite point is not in the South Atlantic region or the bipolar region, if the data jump of two consecutive frames exceeds 50% of the data range of the detection equipment, the next frame will be considered The data are outliers and eliminated; for the proton data at the time of data collection and the position of the sub-satellite point in the South Atlantic region or the bipolar region, if two consecutive frames of data jump exceed 50% of the data range of the detection equipment, and this situation occurs continuously for more than 3 times, the data in this period is considered to be an outlier value and eliminated;

The case where the detection equipment data is heavy ion and electronic data: For the heavy ion and electronic data at the time of data collection and the position of the sub-satellite point is not in the polar region, if the data jump of two consecutive frames exceeds 50% of the data range of the detection equipment, it will be considered as the latter. One frame of data is an outlier and eliminated; for heavy ions and electronic data whose sub-satellite point position is in the polar region at the time of data collection, if two consecutive frames of data jump exceed 50% of the data range of the detection equipment, and If this situation occurs more than 3 times in a row, the data in this period is considered to be an outlier value and eliminated;

Step 12, calculate the high-energy particle flux Lo(T) of the space where the satellite is located in a certain historical period T using the single event effect algorithm, and obtain the quantitative index Go of the SRAM device of the computer SRAM of the satellite star service center in this period T affected by the space radiation environment (T);

The calculation method is:

use the formula Obtain the quantitative index Go(To) affected by the space radiation environment at the time To, where σ(E) is the index of the impact of the particle whose energy spectrum is E on the chip, that is, the number of flips; F(E) is the number of times the energy spectrum is E After the particles are eliminated and preprocessed, they are transformed into the obtained flux Lo(To) at TO time in the T period; δ(E)=cosθ, where θ is the direction of the energy spectrum particle detection channel and the normal direction of the SRAM chip installation The included angle; Go(To) is integrated to obtain Go(T);

In the above calculation method, the energy spectrum of the particle needs to be adaptively selected. If the particle is a proton, the energy spectrum of the proton is selected to be between 10 Mev and 300 Mev; if the particle is a heavy ion, the energy spectrum of the heavy ion is selected from He, C, The energy spectrum of Li, Mg, Ar, Fe family ions; if the particles are electrons, the electronic energy spectrum selects the energy spectrum of all electrons;

Step 13: Use data mining or data statistics methods to obtain the correlation Xo(T) between Go(T) and A(T) at the abnormal occurrence time A(T) of the quantitative indicator Go(T) in the period T ;

Step 2, according to the real-time data of the detection equipment of the satellite-borne high-energy particle flux detection equipment, use the method of step 11 to obtain the high-energy particle flux L(t) of the space where the satellite is located at the current moment t; then according to the high-energy particle flux L(t) utilizes the mode of step 12 to obtain the quantitative index G(t) that the SRAM device of the computer of the satellite star service center is affected by the space radiation environment at time t;

Step 3, for G(t), find A(t) in the relationship Xo(T), and use the Apriori algorithm to obtain multiple supportability values and multiple supportable values according to the relationship between G(t) and A(t). The confidence value corresponding to the degree value, select a relationship with the largest support value among multiple support degree values, if the support degree value is greater than 0.6, the alarm information is "Yes" and the support degree value corresponding to the The reliability value is the early warning value Y(t). If the support value is less than or equal to 0.6, the alarm information is "none".

Beneficial effect:

The method for early warning of frequent abnormal occurrences of on-orbit satellites affected by the space environment based on the anomaly mechanism of the present invention is aimed at the anomaly of autonomous shutdown of the computer of a certain remote sensing satellite star center because of its weak single-event resistance ability of the SRAM device. On the basis of the influence of the single event effect on it, based on the method of correlation research, the early warning of the possible occurrence of the anomaly can be realized by using the data of the spaceborne high-energy particle detection equipment. specific:

1. According to the mechanism of the space radiation environment effect, the impact of the space radiation environment on the satellite is directly related to the high-energy particle flux at the actual position of the satellite. Therefore, this method uses the space-borne high-energy particle detection equipment as the input condition. Environmental information released by environmental research institutions for early warning has the advantage of high credibility;

2. Satellite anomalies affected by the space radiation environment are manifestations of physical logic or physical changes caused by the space environment on satellite equipment and devices, and are directly related to the radiation protection design of the active devices. Therefore, this method is aimed at specific satellites with clear mechanisms Compared with the universal early warning for all abnormalities, it has the advantages of high pertinence and accuracy;

3. This method is based on space environment data, and fundamentally solves the problem that the telemetry parameters associated with the anomaly (the flight status of the computer in the Star Service Center) will change suddenly when the anomaly occurs, and it is difficult to find it through the prediction of the telemetry parameter.

Description of drawings

Fig. 1 is a flow chart of analyzing the relationship between space radiation environment and spacecraft on-orbit abnormality in the present invention;

Fig. 2 is a flow chart of early warning of spacecraft on-orbit abnormality according to the space environment forecast result of the present invention;

Fig. 3 is a schematic diagram of the detection results of each energy channel of the proton carried by the satellite-mounted high-energy particle detector of the present invention;

Fig. 4 is a schematic diagram of the detection results of the high-energy heavy ion>C channel of the satellite-mounted high-energy particle detector of the present invention;

Fig. 5 is a schematic diagram of the comparison between the number of single event flips (G) and satellite abnormal events in the present invention.

Detailed ways

The early warning method of the present invention based on the relationship between satellites in orbit affected by the space environment, which includes:

Step 1, analyze the relationship between particle effects and satellite in-orbit anomalies, and obtain. As shown in Figure 1, the specific content is:

Step 11: Transform the detection equipment data of the space-borne high-energy particle flux detection equipment carried by the satellite into the field using the "telemetry data processing method" after preprocessing to obtain the high-energy particle flux L _O (T);

Some outliers will be generated during the transmission and processing of the data of the satellite-borne high-energy particle flux detection equipment. These outliers will bring bias to the subsequent analysis and calculation, so it is necessary to preprocess the data of the detection equipment to remove the wild. Preprocessing mainly includes three steps:

The first step: Eliminate outliers that exceed the data range of the detection equipment. The data range of the detection equipment is generally 0-5V.

Step 2: Perform data screening according to the type of detection equipment data and the location of the sub-satellite point corresponding to the data collection time:

The case where the detection equipment data is proton data: For the data collection time, the location of the sub-satellite point is not in the South Atlantic region (latitude range 10N-60S, longitude range 20E-100W) or the bipolar region (latitude 70 and above), if a data continuous If the jump of two frames of data exceeds 50% of the data range of the detection equipment, the data of the latter frame is considered to be an outlier and eliminated; for the time of data collection, the position of the sub-satellite point is in the South Atlantic region or the polar region, if a piece of data has two consecutive frames If the data jump exceeds 50% of the data range of the detection equipment, and this situation occurs more than 3 times in a row, the data in this period is considered to be an outlier value and eliminated;

The case where the detection equipment data is heavy ion and electronic data: For the data collection time and the sub-satellite point position is not in the polar region (above latitude 70), if a data jump of two consecutive frames exceeds 50% of the detection equipment data range, The latter frame of data is considered to be an outlier and eliminated; for the case where the sub-satellite point is in the polar region at the time of data collection, if a data jump exceeds 50% of the data range of the detection equipment for two consecutive frames, and this situation occurs continuously If it exceeds 3 times, the data in this period is considered to be an outlier value and eliminated;

Step 3: Set the extraction period of the detection equipment data to be the same as the acquisition period of the telemetry parameters that characterize the satellite and cause the computer of the star affairs center to automatically cut off the machine (for example, set it to 50s).

Step 12, calculate the high-energy particle flux L _O (T) of the space where the satellite is located in a certain period of time T in history, using the single event effect algorithm in the prior art to obtain Quantitative indicator G _O (T) of environmental impact;

According to the satellite anti-space environmental radiation protection measures, the mechanism of single event effects, and the statistical analysis results of on-orbit data, when calculating quantitative indicators, the particle energy spectrum needs to be adaptively selected first. The proton energy spectrum selection: 10 Mev to 300 Mev, heavy ion Spectrum selection He, C, Li, Mg, Ar, Fe family ions, electronic spectrum select electrons of all energy spectrums; and increase detection equipment data installation direction factor δ=cosθ, where θ is the energy spectrum particle detection channel direction and SRAM The included angle of the chip mounting normal direction.

The specific calculation method is as follows:

Step 121 , calculate the indicator of the impact of a single particle in space (energy spectrum E) on the applied chip through the Edmonds function, that is, the number of flips σ(E). The formula is as follows:

σ(E)=σ _SAT exp(-(L _1//e )/E)

Among them, σ _SAT is the saturation cross-section, L _1/e is the energy spectrum value at which the saturation threshold is reduced to 1/e, and the two are the anti-space radiation environment index information of the SRAM device (obtained through ground tests).

Step 122, obtain the probability Go(To) (number of flips) of the single-event effect produced by the spacecraft space full-energy spectrum and all kinds of particles on the target device at this moment, the formula is as follows, wherein F(E) is the particle whose energy spectrum is E in the T period The flux L0(T0) at time TO in .

Go(To)＝ΣF(E)·σ(E)·δ(E)

Point Go (To) to get G _O (T);

Step 13, at the abnormal occurrence time A(T) in this period, use data mining or data statistics to process the quantitative index G _O (T) obtained in step 12 that this period T is affected by the space radiation environment to obtain a credible The correlation relationship X(T) in the time domain between the two whose degrees and correlations exceed the set threshold.

Step 131, discretize and preprocess the quantitative index G _O (T) data, and stitch the preprocessed data together to generate a data analysis matrix. A variety of methods in the prior art can be realized, and only one method is listed here:

(1) Quantitative index G _O (T) Discretization processing GO data: three-level discretization, when the value of GOTO is in the interval [0,100), it is set to 0, in the interval [100,1000), it is set to 1, in the interval [1000,∞) is set to 2. This is a discretization strategy, which is adjusted repeatedly according to the analysis results in actual work.

(2) Case of discretized processing of on-orbit abnormality: Discretize the abnormal occurrence time A (T0) according to the YYYY-MM-DDHH:MM:SS format (year, month, day, hour, minute, and second) The positions of sub-satellite points are discretized into the South Atlantic region, the bipolar region and other regions, and other information is not processed. The vertical axis is abnormal data, and the horizontal axis is GO(t).

Calculation of G(T) data duration impact time index

After analysis, the single event effect will last for a certain period of time, and the relationship between G(T) data and on-orbit anomalies is analyzed by calculating the duration index. The formula for calculating the continuous impact time index is as follows:

$\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; ></span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; m</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}\mathrm{<span\; class="notranslate">\; G</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; e</span>}}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; \&lambda;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span>}$

Among them, E(T) represents the continuous influence time index of G(t) data at time t, T and t are in units of 50 seconds, λ is the attenuation factor, and m is the maximum number of days involved in the calculation. where the values of λ and m are specified by the user. According to the analysis, it can be defined that when λ is 0 and m is 600, the effect is that the sum of G(t) data within 10 minutes is used as the continuous impact time index.

(3) Continuous impact time index of discretized G(T) data

According to the following rules, the continuous impact time index of G(T) data is discretized. When the value of the continuous impact time index E(T) is in the interval [0,600), it is set to 0, and in the interval [600,3000), it is set to 1 , set it to 2 in the interval [3000,6000), set it to 3 in the interval [6000,12000), set it to 4 in the interval [12000,60000), and set it to 5 in the interval [60000, infinity). This is a discretization strategy, which is adjusted repeatedly according to the analysis results in actual work.

Step 132, extracting the association relationship, there are many ways to realize it in the prior art, and only one of them is listed:

The Apriori algorithm is used to analyze the data analysis matrix generated in step 131 to obtain a set of association rules. The format of each association rule is as follows:

{Cause}=>{Fruit} Support Credibility

Among them, "cause" is the discretized description of G(T) data and the continuous impact time index, "result" is the discretized description of the occurrence of the autonomous shutdown of the computer in the Star Affairs Center, and the support and credibility describe the establishment of the causal relationship The probability. Examples are as follows:

{G(t)＝2＝>{Anomaly occurs in the computer of the Star Service Center, which shuts down automatically} 0.6 0.6

Step 2, according to the real-time data of the space-borne high-energy particle flux detection equipment, the early warning of the abnormality of the satellite in orbit is carried out, as shown in Figure 2.

Step 21, according to the real-time data of the spaceborne high-energy particle flux detection equipment carried by the satellite, obtain the high-energy particle flux L(t) in the space where the satellite is located at the current moment (t); same as step 11, and will not be repeated here.

Step 22, according to the high-energy particle flux L(t), use the single event effect algorithm to obtain the quantitative index G(t) of the SRAM device of the satellite star service center computer affected by the space radiation environment at time t; same as step 12, no more details here .

Step 3, corresponding to the relationship between the abnormality and the space radiation environment based on historical data and the relationship between the space environment and the satellite on-orbit anomaly X(T), using the quantitative index G(t), it can be calculated to obtain the abnormality of the computer autonomous shutdown of the satellite star service center at time t Possible early warning information Y(t).

According to the current G(t), select the relationship with the highest credibility in the XO(T) relationship set (credibility must reach 0.6 or more, if not, the alarm message will be "none"). If the support degree is greater than 0.6, Then the support degree is the early warning value Y(t), that is, the possibility of abnormality has reached this value, if there is no support degree greater than 0.6, the alarm information is "none").

Embodiment one

Step 1: Count the number and time of computer shutdowns in the satellite center during a certain period of time in the history of the satellite;

Select satellites between November and the end of December 2009, and count the time when the satellites occurred on the computer of the star service center, as shown in Table 1.

Table 1

serial number Time of occurrence 1. 2009-11-25 16:50 2. 2009-11-25 18:27 3. 2009-11-27 08:05 4. 2009-11-28 06:40 5. 2009-11-28 01:19 6. 2009-12-1 08:44 7. 2009-12-1 10:17 8. 2009-12-5 13:34 9. 2009-12-7 07:08 10. 2009-12-10 14:41 11. 2009-12-12 21:51

12. 2009-12-16 11:13 13. 2009-12-18 22:00 14. 2009-12-19 14:39 15. 2009-12-19 18:47 16. 2009-12-21 16:33 17. 2009-12-25 12:57 18. 2009-12-26 19:34 19. 2009-12-26 23:27 20. 2009-12-27 05:00 twenty one. 2009-12-29 23:20 twenty two. 2009-12-30 23:26 twenty three. 2009-12-31 08:35 twenty four. 2009-12-31 16:18

Step 2 Analyze the high-energy particle flux at the moment when the satellite anomaly occurs through the on-board detection equipment, and calculate the quantitative index; as shown in Figure 3, the data P1-P6 of the six channels of the energy proton flux detector correspond to different energy spectra. As shown in Figure 4, it is GO(T) of this period T.

a) When the spacecraft is abnormal, the high-energy particle flux L _O (T) at this moment is obtained by using the detection equipment data of the space radiation environment detection equipment on board;

b) Using the single event effect algorithm, obtain the quantitative index G _O (T) of the spacecraft affected by the space radiation environment during this period.

Step 3: Extract the relationship between the quantitative index GO(T) and the abnormality of the spacecraft;

Data mining is performed on the anomalies that have occurred in the history of the spacecraft and the quantitative index GO that is affected by the space radiation environment at the time of the anomalies, and the correlation X between the two with a reliability and correlation of more than 80% is obtained.

As shown in Figure 5, the value at the top of the vertical dotted line in Figure 5 indicates the number of single event flips, "+" indicates that there is an abnormality in the corresponding period, and "o" indicates that there is no abnormal event in the corresponding period. Through statistical analysis, it can be seen that the number of flips is greater than At 40.9 units, the chances of the satellite's Star Operations Center computer shutting down is 80%.

Step 4: According to the data of the space-borne high-energy particle detection equipment, an early warning is given to the possible abnormality of the computer's automatic shutdown of the Star Affairs Center.

According to the real-time data of the space-borne high-energy particle flux detection equipment carried by the satellite, the high-energy particle flux L(t) in the space where the satellite is located at the current moment (t) is obtained; using the single event effect algorithm, the SRAM device of the satellite star affairs center computer, in The quantitative index G(t) affected by the space radiation environment at time t. Corresponding to the relationship X(T) between the anomaly and the space radiation environment summarized based on historical data in step 3, the early warning information Y(t) of the possible occurrence of the abnormal shutdown of the computer of the Satellite Star Operations Center at time t can be calculated.

Of course, the present invention can also have other various embodiments, and those skilled in the art can make various corresponding changes and deformations according to the present invention without departing from the spirit and essence of the present invention, but these corresponding Changes and deformations should belong to the scope of protection of the appended claims of the present invention.

Claims (1)

1., based on the method for early warning that the satellite in orbit of incidence relation affects by space environment, it is characterized in that, comprising:

Step 1, to particle effect and satellite in-orbit abnormal incidence relation analyze, obtain the incidence relation Xo (T) that confidence level and the degree of correlation exceed the time domain between the two of setting threshold value:

Step 11, picks wild pre-service to the detecting devices data of the spaceborne high energy particle flux detecting devices of the Seeds of First Post-flight in certain historical period T, is then translated into the high energy particle flux Lo (T) in space residing for this history period T satellite;

Wherein, pick wild pre-service and comprise three steps:

The first step: will the unruly-value rejecting of detecting devices data range be exceeded;

Second step: carry out data screening according to the kind of detecting devices data and the substar position of data acquisition moment corresponding satellite:

Detecting devices data are the situation of proton data: for data acquisition moment, substar position not in the proton data of region, South Atlantic Ocean or two polar regions, if two continuous frames data jump exceedes 50% of detecting devices data range, then after thinking, frame data are outlier and reject; Data acquisition moment, substar position are in the proton data of region, South Atlantic Ocean or two polar regions, if two continuous frames data jump exceedes 50% of detecting devices data range, and this situation occurs more than 3 times continuously, then think that the data of this period are outlier and reject;

Detecting devices data are the situation of heavy ion and electronic data: for heavy ion and the electronic data of data acquisition moment, Bu Liang polar region, substar position, if two continuous frames data jump exceedes 50% of detecting devices data range, then after thinking, frame data are outlier and reject; The heavy ion of two polar regions and electronic data heavy ion and electronic data are in for data acquisition moment substar position, if two continuous frames data jump exceedes 50% of detecting devices data range, and this situation occurs more than 3 times continuously, then think that this period data is outlier and rejects;

Step 12, utilize single particle effect algorithm to calculate to the high energy particle flux Lo (T) in space residing for history period T satellite, obtain the quantizating index Go (T) that satellite Star Service central computer SRAM device affects by space radiation environment at this period T;

Computing method are:

Utilize formula obtain the quantizating index Go (To) that affects by space radiation environment of To moment, wherein, σ (E) for the particle spectrum particle that is E on the index that affects of effect chip namely overturn number of times; F (E) for particle that power spectrum is E carry out rejecting pre-service after be converted into the flux Lo (To) of the acquisition in the TO moment in the T period; δ (E)=cos θ, wherein θ is the angle that this power spectrum particle detection channel direction and sram chip install normal direction; Integration is carried out to Go (To) and obtains Go (T);

Need in above-mentioned computing method to carry out adaptability selection to particle spectrum, if particle is proton, then selection proton spectrum is the power spectrum between 10Mev to 300Mev; If particle is heavy ion, then heavy ion power spectrum selects the power spectrum of He, C, Li, Mg, Ar, Fe race ion; If particle is electronics, then electronic energy spectrum selects the power spectrum of whole electronics;

, there is to the exception of quantizating index Go (T) in this period T the incidence relation Xo (T) that moment A (T) utilizes the method for data mining or data statistics to obtain between Go (T) and A (T) in step 13;

Step 2, according to the detecting devices real time data of the spaceborne high energy particle flux detecting devices of Seeds of First Post-flight, utilizes the mode of step 11 to obtain high energy particle flux L (t) in space residing for current time t satellite; Then according to quantizating index G (t) that high energy particle flux L (t) utilizes the mode of step 12 to obtain satellite Star Service central computer SRAM device to affect by space radiation environment in t;

Step 3, in incidence relation Xo (T), A (t) is found for G (t), adopt Apriori algorithm to obtain according to G (t) and the incidence relation of A (t) and multiplely support angle value and multiple confidence value supporting angle value corresponding, at multiple incidence relation supporting to select in angle value to support that angle value is maximum, if this support angle value is greater than 0.6, then warning message is " having " and confidence value corresponding to this support angle value is early warning value Y (t), if this support angle value is less than or equal to 0.6, then warning message is "None".