CN112803893A - Health state monitoring system of satellite power supply system - Google Patents
Health state monitoring system of satellite power supply system Download PDFInfo
- Publication number
- CN112803893A CN112803893A CN202110079987.0A CN202110079987A CN112803893A CN 112803893 A CN112803893 A CN 112803893A CN 202110079987 A CN202110079987 A CN 202110079987A CN 112803893 A CN112803893 A CN 112803893A
- Authority
- CN
- China
- Prior art keywords
- conversion
- output current
- parameters
- time
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000012544 monitoring process Methods 0.000 title claims abstract description 41
- 230000036541 health Effects 0.000 title claims abstract description 22
- 238000001514 detection method Methods 0.000 claims abstract description 72
- 238000000605 extraction Methods 0.000 claims abstract description 46
- 230000002159 abnormal effect Effects 0.000 claims abstract description 31
- 238000006243 chemical reaction Methods 0.000 claims description 133
- 239000011159 matrix material Substances 0.000 claims description 31
- 238000005259 measurement Methods 0.000 claims description 21
- 238000012549 training Methods 0.000 claims description 21
- 230000001174 ascending effect Effects 0.000 claims description 16
- 230000000630 rising effect Effects 0.000 claims description 12
- 238000007599 discharging Methods 0.000 claims description 6
- 238000005286 illumination Methods 0.000 claims description 4
- 230000009471 action Effects 0.000 claims description 2
- 238000000034 method Methods 0.000 description 54
- 230000005856 abnormality Effects 0.000 description 32
- 230000009466 transformation Effects 0.000 description 23
- 230000008569 process Effects 0.000 description 17
- 238000013139 quantization Methods 0.000 description 15
- 238000010586 diagram Methods 0.000 description 12
- 230000003862 health status Effects 0.000 description 11
- 230000008859 change Effects 0.000 description 9
- 238000004088 simulation Methods 0.000 description 8
- 230000007704 transition Effects 0.000 description 8
- 238000012545 processing Methods 0.000 description 7
- 230000008901 benefit Effects 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 5
- 239000000284 extract Substances 0.000 description 5
- 238000002347 injection Methods 0.000 description 5
- 239000007924 injection Substances 0.000 description 5
- 238000000844 transformation Methods 0.000 description 5
- 230000007423 decrease Effects 0.000 description 4
- 238000002474 experimental method Methods 0.000 description 4
- 238000012360 testing method Methods 0.000 description 4
- 230000002411 adverse Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 230000002123 temporal effect Effects 0.000 description 2
- 230000036962 time dependent Effects 0.000 description 2
- 230000032683 aging Effects 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000006866 deterioration Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000007500 overflow downdraw method Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S19/00—Satellite radio beacon positioning systems; Determining position, velocity or attitude using signals transmitted by such systems
- G01S19/01—Satellite radio beacon positioning systems transmitting time-stamped messages, e.g. GPS [Global Positioning System], GLONASS [Global Orbiting Navigation Satellite System] or GALILEO
- G01S19/13—Receivers
- G01S19/20—Integrity monitoring, fault detection or fault isolation of space segment
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E10/00—Energy generation through renewable energy sources
- Y02E10/50—Photovoltaic [PV] energy
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Computer Security & Cryptography (AREA)
- Computer Networks & Wireless Communication (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Remote Monitoring And Control Of Power-Distribution Networks (AREA)
Abstract
The invention provides a health state monitoring system of a satellite power supply system, which comprises: a first feature extraction and anomaly detection module configured to extract features of respective telemetry parameters according to satellite power system telemetry data characteristics; collecting the master data of each telemetering parameter in real time, and detecting whether the master data is abnormal or not according to the characteristics of the telemetering parameters; the second feature extraction and anomaly detection module is configured to compare the master data and the backup data of each telemetry parameter to form a comparison rule base; and acquiring the master data and the backup data of each telemetering parameter in real time, and detecting whether the master data and/or the backup data are abnormal or not according to a comparison rule base of the telemetering parameters.
Description
Technical Field
The invention relates to the technical field of satellite remote measurement, in particular to a health state monitoring system of a satellite power supply system.
Background
Due to the uncertainty of the space environment of the satellite, the limitation of testing before transmission, component aging and other factors, some abnormal conditions or faults still can occur in the in-orbit operation period of the satellite, and if the abnormal conditions or faults are not found and processed in time, the service life of the satellite is adversely affected, so that the health state of the satellite needs to be monitored in real time by using the telemetering data of the satellite, and the abnormal conditions are processed in time to avoid deterioration. In the traditional satellite health state monitoring process, the satellite telemetry data is downloaded to a ground station through satellite-ground communication, the ground station analyzes and judges the telemetry data, and a corresponding remote control command is sent to remedy after abnormality is found. However, the method has the defects of prolonging in communication and having a measurement and control blind area, and is not beneficial to timely discovering and processing the abnormity or the fault. The on-orbit autonomous processing of the satellite anomaly detection technology also has various problems of low reliability or low efficiency and the like caused by unreasonable algorithm.
Disclosure of Invention
The invention aims to provide a health state monitoring system of a satellite power supply system, which aims to solve the problem that an on-orbit autonomous processing algorithm of the conventional satellite anomaly detection technology is unreasonable.
In order to solve the above technical problem, the present invention provides a health status monitoring system for a satellite power supply system, including:
a first feature extraction and anomaly detection module configured to extract features of respective telemetry parameters according to satellite power system telemetry data characteristics; and
acquiring master data of each telemetering parameter in real time, and detecting whether the master data is abnormal or not according to the characteristics of the telemetering parameters;
the second feature extraction and anomaly detection module is configured to compare the master data and the backup data of each telemetry parameter to form a comparison rule base; and
and acquiring master data and backup data of each telemetering parameter in real time, and detecting whether the master data and/or the backup data are abnormal or not according to a comparison rule base of the telemetering parameters.
Optionally, in the health status monitoring system of the satellite power supply system, the characteristics include a mode conversion relationship, a mode conversion sequence, and a mode longest duration.
Optionally, in the satellite power system health status monitoring system, the telemetry data characteristics of the satellite power system include:
the satellite power system remote measurement parameters comprise solar panel output current and battery pack voltage remote measurement parameters;
the battery pack voltage telemetry parameters are switched between discharging, charging and sleeping in order;
the output current of the solar panel is sequentially converted among a constant current output section, an output current descending section, an output current constant zero section, an output current fast rising section and an output current slow rising section.
Optionally, in the health status monitoring system of the satellite power system,
the satellite is in a stable state in an illumination area, the output current of the solar sailboard is a constant current output section, and the duration time of the constant current output section is the first time;
when the satellite enters a shadow area, the output current of the solar sailboard is an output current descending section, and the duration time of the output current descending section is the second time;
when the satellite is completely in the shadow area, the output current of the solar sailboard is a constant zero section of the output current, and the duration time of the constant zero section of the output current is the third time;
when the satellite leaves the shadow area, the output current of the solar sailboard is an output current fast-rise section, and the duration time of the solar sailboard is the fourth time;
when the satellite completely goes out of the shadow area, the output current of the solar sailboard is an output current slow-rising section, and the duration time of the output current slow-rising section is the fifth time;
the first time is greater than the third time, the third time is greater than the fifth time, the fifth time is greater than the fourth time, and the fourth time is greater than the second time.
Optionally, in the health status monitoring system of the satellite power system,
sequentially converting the battery pack voltage telemetering parameters among discharging, charging and sleeping to obtain a mode conversion relation and a mode conversion sequence of the battery pack voltage telemetering parameters;
according to the sequential conversion of the output current of the solar panel among a constant current output section, an output current descending section, an output current constant zero section, an output current fast rising section and an output current slow rising section, the modal conversion relation and the modal conversion sequence of the output current of the solar panel are obtained.
Optionally, in the satellite power system health state monitoring system, the minimum amplitude of the telemetry parameter of the satellite power system to the maximum amplitude of the telemetry parameter of the satellite power system is evenly divided into multiple orders of magnitude, and each order of magnitude corresponds to one mode;
acquiring satellite power system remote measurement parameters corresponding to a plurality of time points, and corresponding each satellite power system remote measurement parameter to a mode according to the amplitude of each satellite power system remote measurement parameter;
and summing the time points corresponding to the telemetry parameters of all the satellite power systems in one mode to obtain the longest duration time of the modes.
Optionally, in the health status monitoring system of the satellite power supply system, the modality conversion relationship includes first-order modality conversion;
the first-order modality conversion is carried out from the ith modality to the jth modality;
constructing a first-order mode conversion matrix A for each telemetering parameter, wherein the row number of A is equal to the number of modes, and the column number of A is equal to the number of modes;
respectively learning each telemetry parameter, and if the ith modality of the quantized telemetry parameters in the training data can be converted into the jth modality, Aij is 1; otherwise Aij is 0;
the longest duration of a mode is stored on the diagonal of the first order mode conversion matrix a.
Optionally, in the health status monitoring system of the satellite power supply system, the modality conversion relationship includes second-order modality conversion;
the second-order modal conversion includes: an ascending conversion is superimposed with an ascending conversion, an ascending conversion is superimposed with a descending conversion, a descending conversion is superimposed with an ascending conversion, and a descending conversion is superimposed with a descending conversion;
constructing a second-order mode conversion matrix B for each telemetering parameter, wherein the row number of the B is equal to the number of modes, and the column number of the B is 1;
and respectively learning each telemetering parameter, recording the corresponding position of B as a corresponding conversion value if a certain second-order modal conversion occurs to a single telemetering parameter in the training data, and recording the corresponding position of B as 0 if the second-order modal conversion does not occur to the single telemetering parameter.
Optionally, in the health status monitoring system of the satellite power system,
and comparing the measured data of a certain telemetry parameter with a first-order modal conversion matrix A and a second-order modal conversion matrix B, and detecting whether the measured data is abnormal or not if the measured data violates the rule of the first-order modal conversion matrix A or the second-order modal conversion matrix B.
Optionally, in the satellite power system health status monitoring system, the second feature extraction and anomaly detection module performs the following actions:
obtaining a one-to-one correspondence relationship between the master telemetry parameters and the backup telemetry parameters;
subtracting corresponding telemetering parameters in the training data, and counting the maximum difference value after subtraction of each pair of telemetering parameters;
storing the corresponding relation between the maximum difference and the telemetering parameters into a comparison rule base;
acquiring the corresponding relation between the master remote measurement parameter and the backup remote measurement parameter in a comparison rule base;
subtracting the corresponding telemetry parameters to be detected to obtain an actual difference value;
comparing the actual difference value with the corresponding maximum difference value in the comparison rule base;
if the actual difference is smaller than the maximum difference, the primary data and/or the backup data are in a normal condition;
and if the actual difference is larger than the maximum difference, the primary data and/or the backup data are abnormal.
In the health state monitoring system of the satellite power supply system, the characteristics of each telemetering parameter are extracted according to the characteristics of the telemetering data of the satellite power supply system, and whether the master data is abnormal or not is detected according to the characteristics of the telemetering parameters, so that the information contained in the existing telemetering data is deeply mined, and whether the abnormality or not is judged simply by using the distance and the occurrence frequency in the telemetering data; furthermore, after the master data and the backup data of each telemetering parameter are compared, a comparison rule base is formed, whether the master data and/or the backup data are abnormal or not is detected according to the comparison rule base of the telemetering parameters, the synchronization between the master telemetering parameters and the backup telemetering parameters is utilized to improve the feature extraction and abnormality detection algorithm of the first feature extraction and abnormality detection module, the improved algorithm utilizes the time correlation contained in a single telemetering parameter and the correlation with other parameters, and the advantages of the two algorithms are complemented while the reliability is improved. The timely discovery and processing of possible abnormality or fault of the satellite plays an important role in improving the reliability of the satellite and prolonging the service life of the satellite.
The invention verifies the algorithm by using real satellite telemetering data and simulated abnormal data, and the result shows that the improved algorithm has the advantages of low complexity and high reliability.
Drawings
FIG. 1 is a schematic diagram of an exemplary telemetry parameter for a satellite power system in accordance with an embodiment of the invention;
FIG. 2 is a schematic diagram illustrating the output current variation characteristics of a solar panel according to an embodiment of the present invention;
FIG. 3 is a diagram illustrating the maximum duration of each stage after the output current of a normal solar panel is quantized according to an embodiment of the present invention;
FIG. 4 is a schematic diagram illustrating a first-order conversion relationship between the front and rear moments after the output current of the solar panel is quantized according to an embodiment of the present invention;
FIG. 5 is a schematic diagram illustrating a second-order conversion relationship after quantization of the output current of the solar panel according to an embodiment of the present invention;
FIG. 6 is a flow chart of a method for feature extraction and anomaly detection according to an embodiment of the present invention;
FIG. 7 is a flowchart illustrating a method for detecting an anomaly in backup parameters according to an embodiment of the present invention;
FIG. 8 is a flow chart of an improved feature extraction and anomaly detection method according to an embodiment of the present invention;
FIG. 9 is a schematic diagram of a simulated Exception 1 experimental waveform according to an embodiment of the present invention;
FIG. 10 is a schematic diagram of a simulated Exception 2 experimental waveform according to an embodiment of the present invention;
FIG. 11 is a schematic diagram of a simulated anomaly 3 experimental waveform according to an embodiment of the present invention;
FIG. 12 is a diagram illustrating the quantized output current of the solar array A according to an embodiment of the present invention;
FIG. 13 is a schematic diagram of the first order transformation of the mode 94 after quantization of the output current of the solar array A according to an embodiment of the invention;
fig. 14 is a schematic diagram of second-order conversion of the mode 92 after quantization of the output current of the solar panel a according to an embodiment of the invention.
Detailed Description
The health status monitoring system for satellite power supply system according to the present invention will be described in detail with reference to the accompanying drawings and specific embodiments. Advantages and features of the present invention will become apparent from the following description and from the claims. It is to be noted that the drawings are in a very simplified form and are not to precise scale, which is merely for the purpose of facilitating and distinctly claiming the embodiments of the present invention.
Furthermore, features from different embodiments of the invention may be combined with each other, unless otherwise indicated. For example, a feature of the second embodiment may be substituted for a corresponding or functionally equivalent or similar feature of the first embodiment, and the resulting embodiments are likewise within the scope of the disclosure or recitation of the present application.
The invention provides a health state monitoring system of a satellite power supply system, which aims to solve the problem that an on-orbit autonomous processing algorithm of the conventional satellite anomaly detection technology is unreasonable.
The inventor of the invention finds that with the improvement of the processing capability and the storage capability of the satellite-borne computer, the satellite anomaly detection technology is developed towards the direction of on-orbit autonomous processing. The on-satellite autonomous anomaly detection technology can obviously improve the timeliness of fault detection and the autonomy of the satellite. The method which is currently applied is to compare the telemetry parameters with the preset thresholds and rules to complete the anomaly detection, but the thresholds and rules used in the method need to be designed and verified by a great deal of manpower, and if the application object is changed, the thresholds and rules need to be redesigned, so the expansibility of the method is not high.
In recent years, aerospace systems gradually realize that an anomaly detection method based on machine learning has good expansion performance, and research and application of a telemetry data machine learning method for monitoring the health state of a satellite are started, so that methods such as IMS, ocra, OS-SVM and the like are developed. However, the inventor believes that the majority of the spacecraft telemetry data is normal data, and data which is far away from the normal data, has low similarity or low occurrence frequency is abnormal data. Namely, in all of the above methods, only part of information included in the telemetry data, such as distance, frequency of occurrence, etc., is used, and the information included in the existing telemetry data is not deeply mined.
The inventors have found that in practice satellite telemetry data is time-dependent data in which a number of time series are combined.
Aiming at the problems and the insights, the invention provides a feature extraction and anomaly detection algorithm by analyzing the time correlation of the telemetering data in detail, and the algorithm utilizes the information of whether the modes of the telemetering parameters can be converted or not, the conversion sequence, the mode duration and the like to carry out anomaly detection. Under the assumption of all working modes and modal conversion processes of the training data coverage system, the mined time information is saved in a rule base, and when the actual telemetry parameters violate the rules, an exception is considered to occur.
Further, the inventors have found that the above-described feature extraction and anomaly detection methods only exploit the temporal correlation of a single telemetry parameter, while ignoring the correlation with other telemetry parameters. For example, in order to prolong the service life of the satellite, the satellite generally adopts a plurality of backups to improve the reliability, the correlation between backup telemetry parameters is very strong, and the backup telemetry parameters have a consistent change rule, and the performance of the provided feature extraction and anomaly detection method can be improved by utilizing the synchronicity between the backup parameters.
In order to realize the idea, the invention provides a health state monitoring system of a satellite power supply system, which comprises: a first feature extraction and anomaly detection module configured to extract features of respective telemetry parameters according to satellite power system telemetry data characteristics; collecting the master data of each telemetering parameter in real time, and detecting whether the master data is abnormal or not according to the characteristics of the telemetering parameters; the second feature extraction and anomaly detection module is configured to compare the master data and the backup data of each telemetry parameter to form a comparison rule base; and acquiring the master data and the backup data of each telemetering parameter in real time, and detecting whether the master data and/or the backup data are abnormal or not according to a comparison rule base of the telemetering parameters. The characteristics include a modality conversion relationship, a modality conversion sequence, and a longest duration of a modality.
Firstly, the invention analyzes the telemetering data characteristics of the satellite power system, for example, the telemetering parameters of a certain satellite power system are divided into a discrete type and a continuous type: discrete remote measurement parameters such as single machine on-off state flags; continuous telemetry parameters such as current, voltage, and temperature. The invention aims to solve the problem of abnormal detection of continuous data in the telemetering parameters of the power system because the discrete telemetering parameters are stable in state and have fewer values and the abnormal detection is easier. Figure 1 shows the normal behavior of 5 representative telemetry parameters in a satellite power system.
From fig. 1, it can be seen that the voltage and current type continuous telemetry parameters in the power supply system are represented in a form of orderly switching among several modes, and the switching relation of the temperature type telemetry parameters is not obvious. For example, battery pack voltage telemetry parameters are switched between three modes of discharging, charging and sleeping in order; the output current of the solar panel is sequentially switched among five modes, namely constant current output (constant current output section), rapid output current reduction (output current reduction section), almost zero output current (output current constant zero section), rapid output current rise (output current fast rise section) and slow output current rise (output current slow rise section).
Specifically, take the output current of a solar panel in a satellite power system as an example. As shown in fig. 2, the windsurfing board output current is in a steady state at the light irradiation regions t0 to t1 and has a longer duration (first time). When the satellite enters the shadow area t 1-t 2, the current of the sailboard begins to drop rapidly, the current value change at the two moments before and after is large, the duration (second time) of each current state is short, and the drop is directional from high to low. When the satellite is completely in the shadow area t 2-t 3, the windsurfing board current is also in a steady state, but the duration (third time) is obviously shorter than the illumination area. When the satellite is in the earth shadow t 3-t 4, the windsurfing board current begins to rapidly rise from low to high again, and the duration (fourth time) of each current state is also short. And in the period from t4 to t5, the output current of the sailboard slowly rises back to the stable state, and the duration of the output current is the fifth time. The first time is greater than the third time, the third time is greater than the fifth time, the fifth time is greater than the fourth time, and the fourth time is greater than the second time.
Further, orderly converting the battery pack voltage telemetering parameters among discharging, charging and sleeping to obtain a mode conversion relation and a mode conversion sequence of the battery pack voltage telemetering parameters; according to the sequential conversion of the output current of the solar panel among a constant current output section, an output current descending section, an output current constant zero section, an output current fast rising section and an output current slow rising section, the modal conversion relation and the modal conversion sequence of the output current of the solar panel are obtained. Uniformly dividing the minimum amplitude of the telemetry parameter of the satellite power system to the maximum amplitude of the telemetry parameter of the satellite power system into a plurality of orders of magnitude, wherein each order of magnitude corresponds to one mode; acquiring satellite power system remote measurement parameters corresponding to a plurality of time points, and corresponding each satellite power system remote measurement parameter to a mode according to the amplitude of each satellite power system remote measurement parameter; and summing the time points corresponding to the telemetry parameters of all the satellite power systems in one mode to obtain the longest duration time of the modes.
In order to better show the time correlation of the telemetering data, the output current of the solar sailboard is uniformly quantized, and the quantization is 100 orders of magnitude. FIG. 3 is a statistical result of the longest duration of each quantized step number after quantization of normal data for the 10,000 windsurfing boards output currents of FIG. 2. As can be seen from fig. 3, except for the longer duration of the two steady states (illumination and shaded) the duration of the transition between the steady states is short. If the quantization state with shorter duration lasts longer, it is indicated that there may be an abnormality. In view of the above, the present invention may detect an anomaly by detecting whether the duration of a certain quantized value of the actual telemetry parameter exceeds the maximum duration counted by the existing normal training data.
Fig. 4 is a first-order mode conversion relationship diagram of the 10,000 windsurfing boards in fig. 2 at the front and rear moments after the output currents are uniformly quantized. The first-order modality conversion is carried out from the ith modality to the jth modality; in fig. 4, the x-axis is the quantized value of the windsurfing board output current at the previous time, the y-axis is the quantized value of the windsurfing board output current at the subsequent time, the z-axis is the first-order conversion relationship of the quantized values of the windsurfing board output current at the previous and subsequent times, and z has only two values: 0 and 1. If the quantized value of the windsurfing board output current at the previous moment can be converted to the quantized value at the later moment, z is 1, otherwise z is 0. As can be seen from fig. 4, the first-order conversion relationship (i.e., the portion where z is 1) occurring in the normal state is mainly concentrated near the diagonal line. For example, the mode 94 after the windsurfing board current quantization may be converted to the mode 93 or 95 without converting to the mode 92 or other modes farther from the diagonal. This indicates that, in the normal state, the quantized output current value is only converted to the own or adjacent quantization mode, but not to the distant quantization mode. If a transition occurs that is outside the existing normal data representation, it usually means that an anomaly has occurred. In view of the above, the present invention can detect anomalies by detecting whether the telemetry parameters undergo a first-order modal transformation other than that exhibited by the existing training data.
Because the invention quantizes the continuous telemetering parameters, if the quantization interval is selected unreasonably, the telemetering parameters jump back between two adjacent quantization stages and bring difficulty to later abnormal detection, and multi-stage modal conversion is introduced to counteract the adverse effect brought by quantization error. The second-order modal transformation ignores the transformation towards the self and only concerns the variation trend of the telemetering parameters. The second-order modal conversion includes: an ascending conversion is superimposed with an ascending conversion, an ascending conversion is superimposed with a descending conversion, a descending conversion is superimposed with an ascending conversion, and a descending conversion is superimposed with a descending conversion; for example, the relationship 92 → 92 → 92 → 93 of the four adjacent first-order modal transformations is simplified to 92 → 93 in the relationship of multi-order modal transformation, i.e. the telemetry parameters have a rising trend. Each of the multiple-order modality conversion can only be used for other modality conversion. Taking the second-order modal transformation as an example, there are four cases.
1) The former conversion has an ascending trend, and the latter conversion also has an ascending trend;
2) the former conversion has an ascending trend, and the latter conversion has a descending trend;
3) the former conversion has a descending trend, and the latter conversion has an ascending trend;
4) the former conversion has a downward trend, and the latter conversion also has a downward trend.
For example, the second order modal transformation 92 → 93 → 94 follows the first case because the telemetry parameters have a tendency to continuously rise at two adjacent transformations; whereas the second order modal transformation 92 → 93 → 92 corresponds to the second case, since the telemetry parameter has a tendency to fall after having a tendency to rise first in two adjacent transformations. Still make statistics of the second order conversion relationship of the normal data after quantization of the 10,000 windsurfing board output currents in fig. 2, and the result is shown in fig. 5. In fig. 5, the second-order conversion relationship is represented by the Z-axis, and the 4 non-0 values thereof represent the 4 second-order modal conversion conditions. Compared to fig. 4, although the second-order modal transformation that can occur in the windsurfing board output current is also concentrated near the diagonal in the normal state, there are two distinct state transformation processes: one is the continuous descending process of the current of the solar array (the conversion with the second-order conversion relation of 4 in fig. 5 is approximately a straight line); one is the continuous rising process of the current of the solar sailboard (the conversion with the second-order conversion relation of 1 in fig. 5 is also similar to a straight line). As can be seen from fig. 5, the second-order modal transformation relationship can well extract the variation trend of the telemetry parameters, and if the trend represented by the actual telemetry parameters is different from the trend obtained by normal data learning, an abnormality may occur.
To improve reliability or perform certain functions, the critical subsystems and components of the satellite are typically implemented using multiple identical devices or a single machine. For example, physical hot backup and triple modular redundancy can improve the reliability of the satellite, and a plurality of single batteries are connected in series and in parallel to jointly realize the functions of storing and releasing energy, and the same single machine or device forms a backup relation. Under a normal state, the change rule of the telemetry parameters with the backup relation has synchronism. When an exception occurs, the synchronicity of the change regularity between the backup parameters may be destroyed. For example, a power supply system of a certain satellite consists of two solar cell panels, two battery packs, an equalizer and a power supply controller, wherein hardware configurations and remote parameters of the two solar cell panels (A, B) and the two battery packs (A, B) are completely the same, and a backup relationship is formed. In order to show the synchronism of the change rule between the main backup telemetry parameters, four pairs of telemetry parameters with backup relation in a power supply system are selected for comparison, and the average value, the standard deviation and the maximum difference value after the subtraction are respectively counted, wherein the results are shown in table 1.
TABLE 1 backup Change rules of telemetry parameters
As can be seen from table 1, the average value, the standard deviation and the maximum value of the difference obtained by subtracting the four pairs of telemetry parameters with backup relationship are relatively small, and have consistent change rules, so that the purpose of anomaly detection can be achieved by utilizing mutual supervision among the four pairs of telemetry parameters.
The first feature extraction and anomaly detection module is provided with a feature extraction and anomaly detection algorithm which is divided into a learning process and a monitoring process, wherein the learning process comprises the step of extracting the features of all the telemetering parameters according to the characteristics of the telemetering data of the satellite power system, and the monitoring process comprises the step of acquiring the master data of all the telemetering parameters in real time and detecting whether the master data is abnormal or not according to the characteristics of the telemetering parameters.
The invention specifically relates to the time correlation in the satellite telemetry parameters, which is characterized by whether the modes can be converted after the telemetry parameters are quantized, the conversion sequence and trend, the longest duration time of the modes and the like. In order to extract the characteristics, each telemetering parameter is uniformly quantized, and a first-order modal conversion matrix A and a second-order modal conversion matrix B are constructed for each telemetering parameter, wherein the row number of the first-order modal conversion matrix A and the row number of the second-order modal conversion matrix B are both equal to the total number of quantization stages; the number of columns A is equal to the number of modes, and the number of columns B is 1. The square matrix A is used for counting the first-order modal conversion relation and the longest duration of the telemetering parameters, and the square matrix B is used for counting the second-order modal conversion relation of the telemetering parameters. Based on the characteristics, the invention provides a characteristic extraction and anomaly detection algorithm which consists of a learning process and a monitoring process.
The learning process is as follows: and respectively learning each telemetry parameter, and if the ith modality of the quantized telemetry parameters in the training data can be converted into the jth modality, setting Aij to 1 in the matrix A of the telemetry parameters, otherwise, setting Aij to 0. The second-order modal transformation matrix B is obtained by comparing two adjacent modal transformations with the four cases. If one of the four modal conversions occurs to a single telemetry parameter in the training data, adding the corresponding trend of change to the corresponding position of the matrix B; if the above four second-order conversions do not occur in the training data, the value of the corresponding position in the matrix B is set to 0. It should be noted that some second-order modal transformations may have a plurality of transformation trends, and the corresponding position of the corresponding matrix B also has a plurality of values. Furthermore, both matrices a and B are asymmetric, i.e. the modal transformation is directional.
In order to obtain the longest duration of a single telemetry parameter modality, the invention counts the duration of each telemetry parameter quantified modality and then takes the maximum value. The duration is the length of time that elapses from the entry of one modality to the transition to the other modality. According to the difference of the steady-state data and the transient-state data and the discontinuity of the actual telemetry parameters (the satellite is out of the observation interval and the telemetry data cannot be obtained), the state with the longest duration time greater than the threshold value Ts is defined as a steady state, and the state with the longest duration time lower than the threshold value Ts is defined as a transient state. The maximum duration of the steady state is artificially set to infinity, i.e. there is no limit to the duration of the steady state, and the maximum duration of the transition state is the maximum duration that the training data has learned. The acquisition of the longest duration of the telemetry parameter mode can be carried out simultaneously in the learning process of the same order mode conversion matrix A and stored on the diagonal of the first order mode conversion matrix A. After the learning process is completed, the invention obtains the characteristics of the first-order modal conversion relation, the longest modal duration, the second-order modal conversion and the like of each telemetering parameter.
The monitoring process is that the measured data of a certain telemetering parameter is compared with a first-order modal conversion matrix A and a second-order modal conversion matrix B, if the measured data violates the rule of the first-order modal conversion matrix A or the second-order modal conversion matrix B, whether the measured data is abnormal or not is detected; specifically, anomaly detection is performed using two modal transformation matrices a and B for each telemetry parameter that were previously learned. When abnormality detection is carried out, firstly, the telemetering parameters are quantized, then whether the telemetering parameters are converted from the existing mode to other modes is checked, and if conversion does not occur, the duration time of the modes is counted; if a modality conversion has occurred, it is checked whether the conversion occurred is allowed or not, in comparison with the matrices a and B. An anomaly is considered to occur for an actual telemetry parameter if the parameter has undergone a transition that is not allowed by the two modality transition matrices. An abnormality is considered to have occurred if the duration of a single modality exceeds the maximum duration of the modality. The flow chart of the whole feature extraction and anomaly detection method is shown in fig. 6.
The aforementioned feature extraction and anomaly detection algorithms only exploit the temporal correlation contained in a single telemetry parameter and do not fully exploit the synchronous correlation of backup parameters. Therefore, the invention also introduces an improved feature extraction and anomaly detection algorithm (or called backup parameter anomaly detection method) in the second feature extraction and anomaly detection module, namely a method for carrying out anomaly detection by using the synchronism among backup parameters. The difference between the telemetry parameter subtractions between the backups is usually mostly small, but in some special cases (e.g. when the telemetry parameter changes faster) the difference is large. Since the standard deviation is a global statistic, the larger difference values generated under special conditions are averaged, and the standard deviation of the statistical inter-backup telemetry parameter subtracted difference is still smaller. If the difference value obtained by subtracting the main backup telemetering parameters is subjected to anomaly detection by adopting a 3 sigma method, a false alarm is generated when the telemetering parameters change rapidly, so that the method and the device detect anomalies by using the maximum difference value between the backup telemetering parameters.
Specifically, the backup parameter abnormality detection method is also divided into two processes: a learning process and a monitoring process. The learning process comprises three parts, firstly, obtaining the one-to-one corresponding relation of the telemetering parameters between backups, then subtracting the corresponding telemetering parameters in the training data, counting the maximum difference value after subtracting each pair of telemetering parameters, and finally storing the maximum difference value of the difference and the corresponding relation into a comparison rule base. The monitoring process is divided into two parts, firstly, the corresponding relation between backup parameters is obtained in a comparison rule base, then the corresponding telemetering parameters needing to be detected are subtracted and compared with the corresponding subtracted maximum difference value in the comparison rule base, and if the actual difference value is smaller than the maximum difference value, the two parameters are considered to be normal; if the actual difference is greater than the maximum difference, it is determined that an abnormality has occurred in at least one of the two parameters. The whole process is shown in fig. 7 (the arrows in the figure indicate the corresponding relations).
The two methods respectively utilize the time correlation of single telemetering parameters and the synchronism among backup telemetering parameters to detect the abnormity, and can be fused to complement the advantages so as to improve the characteristic extraction and abnormity detection method. The fusion method is as follows, the feature extraction and anomaly detection algorithm only needs to monitor half of the telemetric parameters with backup relationship but not all telemetric parameters, and the backup parameter anomaly detection method detects each pair of telemetric parameters with backup relationship once. The two methods run simultaneously, and if at least one abnormal method detects the abnormality at a certain moment, the moment is considered to be abnormal; otherwise, the moment is normal. As shown in fig. 8, the whole anomaly detection process has two advantages compared with the single method, namely, the improved feature extraction and anomaly detection method which is performed simultaneously after the two methods are fused:
firstly, the improved algorithm only extracts the time correlation of the master telemetering parameters, and compared with the method of using feature extraction and abnormality detection for all telemetering parameters, the method reduces the time and space complexity and saves the precious resources of the spacecraft.
Secondly, the detection results of the two methods form a redundancy relation, and compared with a single abnormality detection algorithm, the reliability of an abnormality detection system can be improved. For example, at a certain time, both methods of different principles determine that the value at that time is normal, and the reliability of the determination result is greatly improved as compared with a case where the value at that time is determined to be normal by one method alone.
In order to verify the feature extraction and anomaly detection method and the improved algorithm effectiveness thereof, the invention carries out a plurality of experiments. Firstly, in order to verify the effectiveness and time complexity of the algorithm, the invention selects 32 (16 pairs) of telemetry parameters with backup relation of a certain satellite power subsystem and divides the parameters into three groups, wherein the first group only uses 8 (4 pairs) of telemetry parameters, the second group uses 16 (8 pairs) of telemetry parameters, and the third group uses all 32 (16 pairs) of telemetry parameters. Training data are data of the parameters from 16 o 'clock in 30.h.9.2017 to 16 o' clock in 30.h.10.2017, 700,350 normal data (with a break in the middle), and test data are data of 429,250 normal data of the parameters from 18 o 'clock 25 min 29 sec in 31.h.10.2017 to 22 min 05 sec in 8 o' clock 15.h.11.2017. The sampling rate of the training data and the test data are sampled every 1 second, and are interpolated and smoothed in advance. The maximum duration threshold Ts is set to 2,000. The feature extraction and anomaly detection method and the improved algorithm thereof adopt uniform quantization, and 100 levels are quantized in total. The simulation environment is thinpadt 460s, the CPU is corei5-6200U, the memory is 8G, and the simulation software is Matlab R2019 b. The original feature extraction and anomaly detection method learns and monitors all telemetry parameters, while the improved feature extraction and anomaly detection method learns and monitors only one half (master) with backup relationship. Because the learning process can be carried out off-line, the invention only counts the time consumption of the monitoring stage, each time consumption monitoring result is an average result after ten times of operation, and the time consumption and the false alarm rate of the three methods are respectively shown in the tables 2 and 3.
TABLE 2 comparison of time consumption for the three methods
TABLE 3 Normal data monitoring results
In order to verify the sensitivity of the feature extraction and anomaly detection method and its improved algorithm of the present invention to anomalies or faults of the power supply system. Different from the previous experiment, the feature extraction and anomaly detection method and the improved algorithm thereof in the subsequent experiment only extract the time correlation of the master remote measurement parameters, and do not extract the features of the modal conversion relation and the like of the backup parameters. The invention has made 3 simulation abnormal experiments. The training data were 650,350 normal data and the test data were 50,000 samples. The method of simulated anomaly and anomaly injection is as follows:
The simulation anomaly 3 simulates an anomaly in which the output current of a single solar panel slowly decays to 98% of the normal value. Since the actual windsurfing board performance decay process is quite slow, it is ideally accelerated to 98% of normal with a fixed slope over a period of 25,000 seconds. The method of the abnormal injection is as follows, 5 ten thousand normal data are taken out from the training data, then the abnormal injection is carried out at 25,000 points, the output current of the solar cell arrays a and B slowly drops to 98% of the normal value at 5 ten thousand points, as shown in fig. 11.
TABLE 4 simulation of fault data anomaly detection results
The abnormality detection results of the 3 simulated abnormalities are shown in table 4. The current quantized values of the solar panel A output current in the stages 24900-25400 are shown in FIG. 12. The quantized value of the solar panel a output current at 24999 is 94, the first order conversion relationship and the longest duration of the mode 94 is shown in fig. 13, and the second order conversion relationship of the mode 92 is shown in fig. 14.
The current of the simulated abnormity 1 at the 25000 point is changed into the original 68/69, the quantified value of the current at the 25000 point is 92, and the mode 94 cannot be directly converted into the mode 92 as can be seen from fig. 13, so that the abnormal mode conversion occurs, and the simulated abnormity 1 is immediately detected by the three methods. In addition, as can be seen from fig. 12, after an abnormality occurs, the output current of the solar panel a fluctuates after being quantized, that is, the output current changes back and forth between the modes 92 and 93, and if there is no second-order mode conversion relationship, a false negative occurs after the mode 92 changes back to the mode 93. In other words, the abnormality cannot be detected at this time with only the first-order modality conversion relationship and the longest duration of the modality. However, as can be seen from the second-order mode conversion relationship of fig. 14, only two cases occur after the solar panel current value is quantized into the mode 92: from mode 92, through two consecutive transitions, up to mode 94; two consecutive transitions from modality 92 fall to modality 90. The second-order modal transformation 92 → 93 → 92 after 25200 in FIG. 12 contradicts the second-order modal transformation relationship of the mode 92 learned from the training data, so there is no missing report in this section after the second-order modal transformation relationship is added.
After the output current of the solar array panel B at the 25000 point of the simulated abnormality 2 is reduced by 2.65A, the quantized current value at the 25000 point is 75, as can be seen from fig. 13, the mode 94 cannot be directly converted to the mode 75, but since the output current of the solar array panel B and the output current of the solar array panel a have a backup relationship, information such as a mode conversion relationship, a longest duration time and the like of the output current of the solar array panel B is not extracted, and therefore the abnormality is not detected by the feature extraction and abnormality detection method, while the abnormality is immediately detected by the backup parameter method by comparing the difference value between the output current of the solar array panel B and the output current of the solar array panel a, further, the abnormality is immediately detected by the improved feature extraction and abnormality detection method, and the false alarm rate are both zero. This means that the improved feature extraction and anomaly detection method does not necessarily extract time-dependent information for all telemetry parameters to detect anomalies.
The output current of the solar panel A starts to slowly decrease at a 25000 point through the simulation abnormity 3, and as can be seen from table 3, the abnormity is not immediately detected by the three methods because the initial current decrease is not obvious, but the abnormity is detected earlier by the feature extraction and abnormity detection method than the backup parameter method along with the expansion of the abnormity. The reason why the abnormality is detected by the feature extraction and abnormality detection method is that the current-quantized mode changes with the slow decrease of the output current, and after the output current decreases from 94 to 93, since the longest duration of the mode 93 is only 329 seconds, the abnormality is detected after 329 seconds which continues to last at the mode 93. The improved feature extraction and anomaly detection method runs both the backup parameter method and the feature extraction and anomaly detection method, so it also detects anomalies at point 33899 and has the lowest false alarm rate. The reason why the improved feature extraction and anomaly detection method has the lowest false alarm rate is that the backup parameter method and the feature extraction and anomaly detection method have certain false alarm omission times, but the false alarm omission times are possibly different, and the improved feature extraction and anomaly detection method detects the anomaly only by detecting the anomaly in the two methods, so the improved feature extraction and anomaly detection method has the lowest false alarm rate.
The invention provides a feature extraction and anomaly detection method by combining two properties of a telemetry parameter of a certain satellite power system, namely time correlation and synchronism of a backup parameter, and improves the method by utilizing the backup parameter. Finally, the actual normal telemetering data and the simulated abnormal data are used for verification, and the result shows that the algorithm and the improved algorithm thereof well utilize the characteristics of the telemetering parameters of the power supply system and have the characteristics of low time complexity and high reliability.
In summary, the above embodiments have described the different configurations of the health status monitoring system of the satellite power system in detail, and it goes without saying that the present invention includes but is not limited to the configurations listed in the above embodiments, and any modifications based on the configurations provided in the above embodiments are within the scope of the present invention. One skilled in the art can take the contents of the above embodiments to take a counter-measure.
The embodiments in the present description are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other. For the system disclosed by the embodiment, the description is relatively simple because the system corresponds to the method disclosed by the embodiment, and the relevant points can be referred to the method part for description.
The above description is only for the purpose of describing the preferred embodiments of the present invention, and is not intended to limit the scope of the present invention, and any variations and modifications made by those skilled in the art based on the above disclosure are within the scope of the appended claims.
Claims (10)
1. A satellite power system state of health monitoring system, comprising:
a first feature extraction and anomaly detection module configured to extract features of respective telemetry parameters according to satellite power system telemetry data characteristics; and
acquiring master data of each telemetering parameter in real time, and detecting whether the master data is abnormal or not according to the characteristics of the telemetering parameters;
the second feature extraction and anomaly detection module is configured to compare the master data and the backup data of each telemetry parameter to form a comparison rule base; and
and acquiring master data and backup data of each telemetering parameter in real time, and detecting whether the master data and/or the backup data are abnormal or not according to a comparison rule base of the telemetering parameters.
2. The satellite power system state of health monitoring system of claim 1, in which the characteristics include modality conversion relationships, modality conversion orders, and modality maximum duration.
3. The satellite power system state of health monitoring system of claim 2, wherein the satellite power system telemetry data characteristics include:
the satellite power system remote measurement parameters comprise solar panel output current and battery pack voltage remote measurement parameters;
the battery pack voltage telemetry parameters are switched between discharging, charging and sleeping in order;
the output current of the solar panel is sequentially converted among a constant current output section, an output current descending section, an output current constant zero section, an output current fast rising section and an output current slow rising section.
4. The satellite power system state of health monitoring system of claim 3,
the satellite is in a stable state in an illumination area, the output current of the solar sailboard is a constant current output section, and the duration time of the constant current output section is the first time;
when the satellite enters a shadow area, the output current of the solar sailboard is an output current descending section, and the duration time of the output current descending section is the second time;
when the satellite is completely in the shadow area, the output current of the solar sailboard is a constant zero section of the output current, and the duration time of the constant zero section of the output current is the third time;
when the satellite leaves the shadow area, the output current of the solar sailboard is an output current fast-rise section, and the duration time of the solar sailboard is the fourth time;
when the satellite completely goes out of the shadow area, the output current of the solar sailboard is an output current slow-rising section, and the duration time of the output current slow-rising section is the fifth time;
the first time is greater than the third time, the third time is greater than the fifth time, the fifth time is greater than the fourth time, and the fourth time is greater than the second time.
5. The satellite power system state of health monitoring system of claim 4,
sequentially converting the battery pack voltage telemetering parameters among discharging, charging and sleeping to obtain a mode conversion relation and a mode conversion sequence of the battery pack voltage telemetering parameters;
according to the sequential conversion of the output current of the solar panel among a constant current output section, an output current descending section, an output current constant zero section, an output current fast rising section and an output current slow rising section, the modal conversion relation and the modal conversion sequence of the output current of the solar panel are obtained.
6. The satellite power system health monitoring system of claim 2, wherein the range from the minimum amplitude of the satellite power system telemetry parameter to the maximum amplitude of the satellite power system telemetry parameter is evenly divided into a plurality of orders of magnitude, each order of magnitude corresponding to a mode;
acquiring satellite power system remote measurement parameters corresponding to a plurality of time points, and corresponding each satellite power system remote measurement parameter to a mode according to the amplitude of each satellite power system remote measurement parameter;
and summing the time points corresponding to the telemetry parameters of all the satellite power systems in one mode to obtain the longest duration time of the modes.
7. The satellite power system state of health monitoring system of claim 6, in which the modality conversion relationship comprises a first order modality conversion;
the first-order modality conversion is carried out from the ith modality to the jth modality;
constructing a first-order mode conversion matrix A for each telemetering parameter, wherein the row number of A is equal to the number of modes, and the column number of A is equal to the number of modes;
respectively learning each telemetry parameter, and if the ith modality of the quantized telemetry parameters in the training data can be converted into the jth modality, Aij is 1; otherwise Aij is 0;
the longest duration of a mode is stored on the diagonal of the first order mode conversion matrix a.
8. The satellite power system state of health monitoring system of claim 7, in which the modal conversion relationship comprises a second order modal conversion;
the second-order modal conversion includes: an ascending conversion is superimposed with an ascending conversion, an ascending conversion is superimposed with a descending conversion, a descending conversion is superimposed with an ascending conversion, and a descending conversion is superimposed with a descending conversion;
constructing a second-order mode conversion matrix B for each telemetering parameter, wherein the row number of the B is equal to the number of modes, and the column number of the B is 1;
and respectively learning each telemetering parameter, recording the corresponding position of B as a corresponding conversion value if a certain second-order modal conversion occurs to a single telemetering parameter in the training data, and recording the corresponding position of B as 0 if the second-order modal conversion does not occur to the single telemetering parameter.
9. The satellite power system state of health monitoring system of claim 8,
and comparing the measured data of a certain telemetry parameter with a first-order modal conversion matrix A and a second-order modal conversion matrix B, and detecting whether the measured data is abnormal or not if the measured data violates the rule of the first-order modal conversion matrix A or the second-order modal conversion matrix B.
10. The satellite power system state of health monitoring system of claim 1, wherein the second feature extraction and anomaly detection module performs the following actions:
obtaining a one-to-one correspondence relationship between the master telemetry parameters and the backup telemetry parameters;
subtracting corresponding telemetering parameters in the training data, and counting the maximum difference value after subtraction of each pair of telemetering parameters;
storing the corresponding relation between the maximum difference and the telemetering parameters into a comparison rule base;
acquiring the corresponding relation between the master remote measurement parameter and the backup remote measurement parameter in a comparison rule base;
subtracting the corresponding telemetry parameters to be detected to obtain an actual difference value;
comparing the actual difference value with the corresponding maximum difference value in the comparison rule base;
if the actual difference is smaller than the maximum difference, the primary data and/or the backup data are in a normal condition;
and if the actual difference is larger than the maximum difference, the primary data and/or the backup data are abnormal.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110913372.3A CN113589330B (en) | 2021-01-21 | 2021-01-21 | Satellite power supply system health state monitoring system |
| CN202110079987.0A CN112803893B (en) | 2021-01-21 | 2021-01-21 | Health state monitoring system of satellite power supply system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110079987.0A CN112803893B (en) | 2021-01-21 | 2021-01-21 | Health state monitoring system of satellite power supply system |
Related Child Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110913372.3A Division CN113589330B (en) | 2021-01-21 | 2021-01-21 | Satellite power supply system health state monitoring system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112803893A true CN112803893A (en) | 2021-05-14 |
| CN112803893B CN112803893B (en) | 2021-08-10 |
Family
ID=75810958
Family Applications (2)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110079987.0A Active CN112803893B (en) | 2021-01-21 | 2021-01-21 | Health state monitoring system of satellite power supply system |
| CN202110913372.3A Active CN113589330B (en) | 2021-01-21 | 2021-01-21 | Satellite power supply system health state monitoring system |
Family Applications After (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110913372.3A Active CN113589330B (en) | 2021-01-21 | 2021-01-21 | Satellite power supply system health state monitoring system |
Country Status (1)
| Country | Link |
|---|---|
| CN (2) | CN112803893B (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20070057649A (en) * | 2005-12-01 | 2007-06-07 | 한국전자통신연구원 | System for managing falut using the technology of satellite telemetering and its method |
| CN109934130A (en) * | 2019-02-28 | 2019-06-25 | 中国空间技术研究院 | The in-orbit real-time fault diagnosis method of satellite failure and system based on deep learning |
| CN110348150A (en) * | 2019-07-17 | 2019-10-18 | 上海微小卫星工程中心 | A kind of fault detection method based on dependent probability model |
| CN110909822A (en) * | 2019-12-03 | 2020-03-24 | 中国科学院微小卫星创新研究院 | Satellite anomaly detection method based on improved Gaussian process regression model |
| CN111400865A (en) * | 2020-02-18 | 2020-07-10 | 北京空间飞行器总体设计部 | Satellite single-machine component health assessment method based on slowly-varying analog quantity telemetry data |
| CN111474491A (en) * | 2020-03-10 | 2020-07-31 | 上海卫星工程研究所 | Satellite power supply abnormity detection method based on pattern analysis |
| CN111680748A (en) * | 2020-06-08 | 2020-09-18 | 中国人民解放军63920部队 | Identification method and identification device for spacecraft state mode |
| CN111874268A (en) * | 2020-07-22 | 2020-11-03 | 中国科学院微小卫星创新研究院 | Satellite centralized type autonomous health management system |
| CN111917453A (en) * | 2020-07-15 | 2020-11-10 | 中国科学院微小卫星创新研究院 | Satellite integrated electronic system |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104297699B (en) * | 2014-09-26 | 2017-03-15 | 航天东方红卫星有限公司 | A kind of satellite power supply health status tracking and detection method based on intelligent interpretation |
| RU2611568C1 (en) * | 2016-01-12 | 2017-02-28 | Российская Федерация, от имени которой выступает Федеральное космическое агентство | Method of nickel-hydrogen accumulator batteries operation in power supply system of spacecraft with large service life |
| CN108734360B (en) * | 2018-06-19 | 2021-09-07 | 哈尔滨工业大学 | Correction-based ELM prediction model multidimensional telemetering data intelligent interpretation method |
| CN111193474B (en) * | 2020-01-14 | 2020-11-20 | 北京空间飞行器总体设计部 | High-precision autonomous diagnosis method for satellite solar wing output current |
-
2021
- 2021-01-21 CN CN202110079987.0A patent/CN112803893B/en active Active
- 2021-01-21 CN CN202110913372.3A patent/CN113589330B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20070057649A (en) * | 2005-12-01 | 2007-06-07 | 한국전자통신연구원 | System for managing falut using the technology of satellite telemetering and its method |
| CN109934130A (en) * | 2019-02-28 | 2019-06-25 | 中国空间技术研究院 | The in-orbit real-time fault diagnosis method of satellite failure and system based on deep learning |
| CN110348150A (en) * | 2019-07-17 | 2019-10-18 | 上海微小卫星工程中心 | A kind of fault detection method based on dependent probability model |
| CN110909822A (en) * | 2019-12-03 | 2020-03-24 | 中国科学院微小卫星创新研究院 | Satellite anomaly detection method based on improved Gaussian process regression model |
| CN111400865A (en) * | 2020-02-18 | 2020-07-10 | 北京空间飞行器总体设计部 | Satellite single-machine component health assessment method based on slowly-varying analog quantity telemetry data |
| CN111474491A (en) * | 2020-03-10 | 2020-07-31 | 上海卫星工程研究所 | Satellite power supply abnormity detection method based on pattern analysis |
| CN111680748A (en) * | 2020-06-08 | 2020-09-18 | 中国人民解放军63920部队 | Identification method and identification device for spacecraft state mode |
| CN111917453A (en) * | 2020-07-15 | 2020-11-10 | 中国科学院微小卫星创新研究院 | Satellite integrated electronic system |
| CN111874268A (en) * | 2020-07-22 | 2020-11-03 | 中国科学院微小卫星创新研究院 | Satellite centralized type autonomous health management system |
Non-Patent Citations (2)
| Title |
|---|
| GE ZHANG 等: "A Satellite Incipient Fault Detection Method Based on Local Optimum Projection Vector and Kullback-Leibler Divergence", 《APPLIED SCIENCE》 * |
| 张怀峰 等: "面向卫星电源系统的一种新颖异常检测方法", 《宇航学报》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112803893B (en) | 2021-08-10 |
| CN113589330A (en) | 2021-11-02 |
| CN113589330B (en) | 2024-04-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110221198B (en) | Coal mine underground combination switch fault diagnosis method based on fault tree | |
| CN106772080B (en) | Time equivalence modeling method for accelerated degradation test of space lithium ion battery | |
| CN112288326B (en) | Fault scene set reduction method suitable for toughness evaluation of power transmission system | |
| Fesharaki et al. | Simultaneous optimal design of measurement and communication infrastructures in hierarchical structured WAMS | |
| Lin et al. | An optimization framework for dynamically reconfigurable battery systems | |
| CN114236393B (en) | Method and system for detecting battery abnormality on line based on big data | |
| CN110137944B (en) | Voltage stability disturbance source positioning method based on random matrix theory | |
| Wang et al. | Fractal characteristics analysis of blackouts in interconnected power grid | |
| US11709486B2 (en) | Fault signal recovery system and method | |
| CN110646708A (en) | 10kV single-core cable early state identification method based on double-layer long-and-short-term memory network | |
| CN112803893B (en) | Health state monitoring system of satellite power supply system | |
| Wei et al. | Dynamic modeling and resilience for power distribution | |
| CN116401532A (en) | Method and system for recognizing frequency instability of power system after disturbance | |
| CN110807172A (en) | Risk prediction method and device for relay protection of power system | |
| Dobson et al. | Towards quantifying cascading blackout risk | |
| CN108228800B (en) | Photovoltaic power generation system anomaly detection system and method based on data mining | |
| Angelov et al. | Investigation of the Reliability of Service Equipment in Communication Networks | |
| Herzog et al. | Dynamics sensor validation for reusable launch vehicle propulsion | |
| Zhan et al. | Fault location method of distribution network based on FTU information comparison and search | |
| CN117434439B (en) | Circuit and method for collecting state of large-scale device with auxiliary contact | |
| Singh et al. | A new method to perform lithium-ion battery pack fault diagnostics–Part 3: Adaptation for fast charging | |
| CN116150233B (en) | Minimum mapping processing method and system for active power distribution station operation state data | |
| CN113949343B (en) | Method and device for detecting photovoltaic power generation equipment, electronic equipment and storage medium | |
| Matthews et al. | Simulating the stochastic dynamics and cascade failure of power networks | |
| Guo et al. | Detection of Transient Signals in Smart Grid Using Artificial Neural Network Modeling and JAYA Optimization |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |