Spacecraft control system fault identification method and device based on multi-source information fusion

Abstract

The invention discloses a spacecraft control system fault identification method and device based on multi-source information fusion. The method comprises the following steps: performing multi-source fusion on data of a spacecraft in development and test processes and on-orbit historical data to obtain a plurality of fault samples; each fault sample is provided with a fault label, and each fault label is used for representing a fault mode corresponding to the fault sample; normalizing each fault sample to obtain a normalized fault sample; clustering the normalized fault sample based on a preset clustering algorithm to obtain a plurality of clusters and a fault mode of each cluster; each cluster at least comprises one core object; for the on-orbit data at any moment, calculating the distance between the on-orbit data and the core object in each cluster so as to determine the cluster to which the on-orbit data belongs based on the calculated distance; and determining the fault mode corresponding to the cluster to which the on-orbit data belongs as an identification result. According to the invention, the accuracy of spacecraft fault identification can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of fault diagnosis, and in particular to a method and device for identifying faults in a spacecraft control system based on multi-source information fusion. Background Art

[0002] At present, when performing fault identification of spacecraft control systems, whether it is on-board autonomous identification or ground-assisted identification, model-based methods or signal processing methods are mainly used. These two methods generally identify faults based on the appearance of on-orbit faults, without considering the data of product development and testing processes and the fault knowledge contained in the on-orbit historical data, resulting in a waste of useful information.

[0003] Therefore, there is an urgent need for a spacecraft control system fault identification method and device based on multi-source information fusion to solve the above problems. Summary of the invention

[0004] The present invention provides a method and device for identifying faults in a spacecraft control system based on multi-source information fusion, which can improve the accuracy of identifying faults in a spacecraft. The technical solution is as follows:

[0005] On the one hand, a method for identifying faults in a spacecraft control system based on multi-source information fusion is provided, the method comprising:

[0006] The data of the spacecraft during the development and testing process and the historical data on orbit are multi-source fused to obtain multiple fault samples; each of the fault samples is provided with a fault label, and each of the fault labels is used to characterize the fault mode corresponding to the fault sample;

[0007] Performing normalization processing on each of the fault samples to obtain a normalized fault sample;

[0008] Based on a preset clustering algorithm, the normalized fault samples are clustered to obtain multiple clusters and the fault mode of each cluster; wherein each cluster includes at least one core object;

[0009] For the on-orbit data of the spacecraft at any moment, the distance between the on-orbit data and the core object in each cluster is calculated, so as to determine the cluster to which the on-orbit data belongs based on the calculated distance;

[0010] The failure mode corresponding to the cluster to which the on-track data belongs is determined as the identification result of the on-track data.

[0011] On the other hand, a spacecraft control system fault identification device based on multi-source information fusion is provided, the device comprising:

[0012] A fusion unit is used to fuse the data of the spacecraft during the development and testing process and the historical data on orbit to obtain multiple fault samples; each of the fault samples is provided with a fault label, and each of the fault labels is used to characterize the fault mode corresponding to the fault sample;

[0013] A processing unit, used for performing normalization processing on each of the fault samples to obtain a normalized fault sample;

[0014] A clustering unit, used for clustering the normalized fault samples based on a preset clustering algorithm to obtain multiple clusters and a fault mode of each cluster; wherein each cluster includes at least one core object;

[0015] a calculation unit, configured to calculate, for the on-orbit data of the spacecraft at any moment, a distance between the on-orbit data and a core object in each cluster, so as to determine the cluster to which the on-orbit data belongs based on the calculated distance;

[0016] The determination unit is used to determine the failure mode corresponding to the cluster to which the on-track data belongs as the identification result of the on-track data.

[0017] On the other hand, a computer-readable storage medium is provided, wherein a computer program is stored in the storage medium, and when the computer program is executed by a processor, the steps of the above-mentioned method for fault identification of spacecraft control system based on multi-source information fusion are implemented.

[0018] On the other hand, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the steps of the above-mentioned method for fault identification of a spacecraft control system based on multi-source information fusion.

[0019] The embodiment of the present invention provides a method for fault identification of a spacecraft control system based on multi-source information fusion. The method fully considers the data of the spacecraft during the development and testing process and the historical data on orbit. Through multi-source information fusion, the fault samples determined contain more useful information. Therefore, by clustering the fault samples, the obtained clusters and the fault mode of each cluster are more accurate. Finally, based on the obtained clusters, the fault identification of the on-orbit data is performed, and the obtained results are more accurate. It can be seen that the present application can improve the accuracy of spacecraft fault identification. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0021] Figure 1 It is a flow chart of a method for identifying faults in a spacecraft control system based on multi-source information fusion provided by an embodiment of the present invention;

[0022] Figure 2 It is a structural diagram of a spacecraft control system fault identification device based on multi-source information fusion provided by an embodiment of the present invention;

[0023] Figure 3 It is a hardware architecture diagram of a computer device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0024] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0025] The specific implementation of the above concept is described below.

[0026] Please refer to Figure 1 , an embodiment of the present invention provides a method for identifying faults in a spacecraft control system based on multi-source information fusion, the method comprising:

[0027] Step 100, multi-source fusion of data from the spacecraft during development and testing and historical data on orbit is performed to obtain multiple fault samples; each fault sample is provided with a fault label, and each fault label is used to characterize the fault mode corresponding to the fault sample;

[0028] Step 102, normalizing each fault sample to obtain a normalized fault sample;

[0029] Step 104, clustering the normalized fault samples based on a preset clustering algorithm to obtain multiple clusters and a fault mode of each cluster; wherein each cluster includes at least one core object;

[0030] Step 106, for the spacecraft on-orbit data at any moment, calculate the distance between the on-orbit data and the core object in each cluster, so as to determine the cluster to which the on-orbit data belongs based on the calculated distance;

[0031] Step 108: Determine the fault mode corresponding to the cluster to which the on-track data belongs as the identification result of the on-track data.

[0032] In this embodiment, the data of the spacecraft during the development and testing process and the historical data on orbit are fully considered. Through multi-source information fusion, the fault samples determined contain more useful information. Therefore, by clustering the fault samples, the obtained clusters and the fault mode of each cluster are more accurate. Finally, based on the obtained clusters, the fault identification of the on-orbit data is performed, and the obtained results are more accurate. It can be seen that the present application can improve the accuracy of spacecraft fault identification.

[0033] Described below Figure 1 How the various steps are performed.

[0034] First, regarding step 100 and step 102:

[0035] In the above steps, the fault label is used to characterize the failure mode corresponding to the fault sample. For example, if the fault label of a fault sample is 1, it means that its failure mode is failure mode 1; fault label 2 represents failure mode 2, and so on. Each failure mode represents a different hazard level. For example, some faults are single-point failures. Once they occur, they will have a serious impact on the spacecraft, while some faults will have little impact on the spacecraft. Therefore, when performing fault identification, it is necessary to fully consider the fault hazard level and focus on the failure of key components. Usually, the hazard level is closely related to the severity of the fault, whether it is a single-point failure, the probability of failure, the component to which the fault belongs, etc., and is generally determined by the spacecraft FMEA.

[0036] In addition, due to the differences in the formats of data collected during the development and testing of spacecraft and historical data on orbit, it is necessary to standardize the data to facilitate subsequent data analysis.

[0037] For step 104, the preset clustering algorithm is the DBSCAN algorithm. Based on the preset clustering algorithm, the normalized fault samples are clustered to obtain multiple clusters and the fault mode of each cluster, including:

[0038] The DBSCAN algorithm is optimized based on the particle swarm algorithm to obtain the optimal domain radius and the optimal minimum number of neighbors of the DBSCAN algorithm;

[0039] Take each fault sample as the target fault sample in turn, and determine a circular area with the target fault sample as the center and the optimal domain radius as the radius; traverse other fault samples, and for each fault sample traversed, calculate the Euclidean distance between the fault sample and the target fault sample, and determine whether the calculated Euclidean distance is less than the optimal domain radius; if so, determine that the fault sample is within the circular area, if not, determine that the fault sample is not within the circular area; until each fault sample is traversed, the number of fault samples within the circular area is obtained;

[0040] Determine whether the number of fault samples in the circular area is greater than the optimal minimum neighbor number; if so, determine that the target fault sample is a core object, and determine all density-connected fault samples starting from the core object;

[0041] For each determined core object, the following steps are performed: traverse all core objects within the optimal domain radius of the core object, and find all fault samples that are densely connected to these core objects to form a corresponding cluster, and determine the fault mode of the cluster based on the fault label of each fault sample contained in the cluster.

[0042] In this step, the DBSCAN (Density-Based Spatial Clustering of Application with Noise) algorithm can divide the areas with sufficiently high density into clusters according to the data distribution density, thereby realizing fault identification by determining the cluster where the spacecraft on-orbit data is located.

[0043] In some implementations, determining the failure mode of the cluster based on the failure label of each failure sample included in the cluster includes:

[0044] Count the number of fault samples corresponding to each fault label in the cluster respectively;

[0045] The fault mode represented by the fault label with the largest number of corresponding fault samples is determined as the fault mode of the cluster.

[0046] For example, a cluster contains 20 fault samples, of which 15 have fault labels of fault label 2 and 5 have fault labels of fault label 1. Since the number of fault samples corresponding to fault label 2 in this cluster is the largest, the fault mode represented by fault label 2 is taken as the fault mode of this cluster.

[0047] In some implementations, the preset clustering algorithm is optimized based on the particle swarm algorithm to obtain the optimal domain radius and the optimal minimum neighbor number of the clustering algorithm, including:

[0048] S1, initialize the particle swarm parameters, which include the number of particles, the initial best position of the particle swarm, the initial position of each particle, the initial speed and the initial best position of the particle;

[0049] S2, traverse each particle in the particle swarm in turn, and for each traversed particle, execute the following: take each fault sample as a sub-target fault sample in turn, and determine a circular area with the sub-target fault sample as the center and the domain radius of the current cycle as the radius; traverse other fault samples, and for each traversed fault sample, calculate the Euclidean distance between the fault sample and the sub-target fault sample, and determine whether the calculated Euclidean distance is less than the domain radius of the current cycle; if so, determine that the fault sample is within the circular area, if not, determine that the fault sample is not within the circular area; until each fault sample is traversed, the number of fault samples in the circular area is obtained; determine whether the number of fault samples in the circular area is greater than the minimum number of neighbors of the current cycle; if so, determine that the sub-target fault sample is a core object, and determine all density-connected fault samples starting from the core object;

[0050] For each determined core object, the following steps are performed: traversing all core objects within the domain radius of the current loop of the core object, and finding all fault samples that are densely connected to these core objects to form a corresponding cluster, and determining the fault mode of the cluster based on the fault label of each fault sample contained in the cluster;

[0051] For each determined cluster, the following steps are performed: traverse each fault sample in the cluster and determine whether the fault mode corresponding to the fault label of the traversed fault sample is consistent with the fault mode of the cluster; if so, retain the fault sample; if not, remove the fault sample from the cluster; until each fault sample in the cluster is traversed and the accurate number of samples in the cluster is obtained;

[0052] S3, based on the number of accurate samples in each cluster, calculate the fault recognition accuracy of the current cycle, and execute S4;

[0053] S4, judging whether the fault recognition accuracy of the current cycle is greater than the accuracy corresponding to the current best position of the current particle, if so, updating the best position of the current particle; judging whether the fault recognition accuracy of the current cycle is greater than the accuracy corresponding to the current cycle best position of the current particle group, if so, updating the best position of the current particle group;

[0054] S5, determine whether the preset stop condition is met; if not, update the position and speed of each particle, increase the number of iterations by 1, and execute S2 to S4 repeatedly until the preset stop condition is met to obtain the final best position of the particle swarm;

[0055] S6, taking the domain radius and the minimum neighbor number corresponding to the final best position of the particle swarm as the optimal domain radius and the optimal minimum neighbor number of the clustering algorithm.

[0056] In this embodiment, assuming that the size of the particle group is n, the position of the i-th particle is represented by x id =[x i1 x i2 ], where x i1 =Eps,x i2 =MinPts, i=1,2,...n, d=1,2. Its speed is expressed as v id The best position it has ever experienced is denoted as pid, and the best position the particle swarm has ever experienced is denoted as p gd ; Eps is the domain radius of the DBSCAN algorithm; MinPts is the minimum number of neighbors of the DBSCAN algorithm.

[0057] Then the i-th particle updates its speed and position according to the following formula:

[0058] v id =ωv id +c1r1(p id -x id )+c2r2(p gd -x id )

[0059] x id =x id +v id

[0060] Where i = 1, 2, ... n, d = 1, 2; ω, c1 and c2 are constants, usually ω = 1, c1 = c2 = 2, γ1 and γ2 are random numbers between [0, 1], x id and v id The initial value of is usually given randomly.

[0061] It should be noted that this embodiment fully considers the fault identification accuracy of the fault hazard level when performing parameter optimization, which can improve the accuracy of the optimization result.

[0062] It should also be noted that the following method is used to determine whether two parameters are density-connected:

[0063] Assuming p and q are any two points in the data set, density reachability mainly means that there is at least one data point k that satisfies: p∈N eps (k),q∈N eps (k), where N eps (k) represents the set of neighboring points with k as the center and Eps as the neighborhood radius. If p and q can reach the same point, then p and q are said to be density connected.

[0064] In some implementations, the fault recognition rate is calculated using the following formula:

[0065]

[0066] Where c is the fault recognition rate; j = 1, 2...N, N is the number of fault modes; w j N is the hazard level of the jth failure mode; j is the total number of fault samples contained in the cluster corresponding to the jth fault mode; is the number of correct fault samples contained in the cluster corresponding to the jth fault mode.

[0067] In this step, the hazard level is closely related to the severity of the fault, whether it is a single point fault, the probability of the fault occurring, and the component to which the fault belongs, and is generally determined by the spacecraft FMEA.

[0068] In some embodiments, the preset stop condition is:

[0069] The accuracy corresponding to the best position of the current cycle of the particle swarm is greater than the accuracy threshold;

[0070] And / or, the number of iterations is greater than a threshold number.

[0071] Finally, for steps 106 and 108:

[0072] Step 106 includes: traversing each cluster, and for each traversed cluster, executing:

[0073] Calculate the distance between the on-track data and each core object in the cluster; and determine whether there is a core object that satisfies the distance from the on-track data to the core object is less than the optimal domain radius; if so, determine that the on-track data belongs to the cluster; if not, traverse the next cluster until the cluster to which the track data belongs is determined.

[0074] After the cluster to which the on-track data belongs is determined, the failure mode corresponding to the cluster is the failure mode of the on-track data, that is, the identification result. For example, if the failure mode of the cluster is a single point failure, then the failure mode of the on-track data is a single point failure.

[0075] like Figure 2 , Figure 3 As shown, an embodiment of the present invention provides a spacecraft control system fault identification device based on multi-source information fusion. The device embodiment can be implemented by software, or by hardware or a combination of software and hardware. From the hardware level, Figure 2 As shown in FIG. 1 , a hardware architecture diagram of a computing device where a spacecraft control system fault identification device based on multi-source information fusion is provided in an embodiment of the present invention is located. Figure 2In addition to the processor, memory, network interface, and non-volatile memory shown in the figure, the computing device in which the device is located in the embodiment may also generally include other hardware, such as a forwarding chip responsible for processing messages, etc. Taking software implementation as an example, Figure 3 As shown, as a device in a logical sense, the CPU of the computing device in which it is located reads the corresponding computer program in the non-volatile memory into the internal memory and runs it.

[0076] Please refer to Figure 3 The embodiment of the present invention provides a spacecraft control system fault identification device based on multi-source information fusion, the device comprising:

[0077] The fusion unit 300 is used to fuse the data of the spacecraft during the development and testing process and the historical data on orbit to obtain multiple fault samples; each fault sample is provided with a fault label, and each fault label is used to characterize the fault mode corresponding to the fault sample;

[0078] The processing unit 302 is used to perform normalization processing on each fault sample to obtain a normalized fault sample;

[0079] The clustering unit 304 is used to perform clustering processing on the normalized fault samples based on a preset clustering algorithm to obtain multiple clusters and a fault mode of each cluster; wherein each cluster includes at least one core object;

[0080] A calculation unit 306 is used to calculate the distance between the on-orbit data and the core object in each cluster for the on-orbit data of the spacecraft at any time, so as to determine the cluster to which the on-orbit data belongs based on the calculated distance;

[0081] The determination unit 308 is configured to determine the failure mode corresponding to the cluster to which the on-track data belongs as the identification result of the on-track data.

[0082] In some embodiments, the preset clustering algorithm is a DBSCAN algorithm;

[0083] The clustering unit 304 is used to perform the following operations:

[0084] The DBSCAN algorithm is optimized based on the particle swarm algorithm to obtain the optimal domain radius and the optimal minimum number of neighbors of the DBSCAN algorithm;

[0085] Take each fault sample as the target fault sample in turn, and determine a circular area with the target fault sample as the center and the optimal domain radius as the radius; traverse other fault samples, and for each fault sample traversed, calculate the Euclidean distance between the fault sample and the target fault sample, and determine whether the calculated Euclidean distance is less than the optimal domain radius; if so, determine that the fault sample is within the circular area, if not, determine that the fault sample is not within the circular area; until each fault sample is traversed, the number of fault samples within the circular area is obtained;

[0086] Determine whether the number of fault samples in the circular area is greater than the optimal minimum neighbor number; if so, determine that the target fault sample is a core object, and determine all density-connected fault samples starting from the core object;

[0087] For each determined core object, the following steps are performed: traverse all core objects within the optimal domain radius of the core object, and find all fault samples that are densely connected to these core objects to form a corresponding cluster, and determine the fault mode of the cluster based on the fault label of each fault sample contained in the cluster.

[0088] In some implementations, determining the failure mode of the cluster based on the failure label of each failure sample included in the cluster includes:

[0089] Count the number of fault samples corresponding to each fault label in the cluster respectively;

[0090] The fault mode represented by the fault label with the largest number of corresponding fault samples is determined as the fault mode of the cluster.

[0091] In some implementations, the preset clustering algorithm is optimized based on the particle swarm algorithm to obtain the optimal domain radius and the optimal minimum neighbor number of the clustering algorithm, including:

[0092] S1, initialize the particle swarm parameters, which include the number of particles, the initial best position of the particle swarm, the initial position of each particle, the initial speed and the initial best position of the particle;

[0093] S2, traverse each particle in the particle swarm in turn, and for each traversed particle, execute the following: take each fault sample as a sub-target fault sample in turn, and determine a circular area with the sub-target fault sample as the center and the domain radius of the current cycle as the radius; traverse other fault samples, and for each traversed fault sample, calculate the Euclidean distance between the fault sample and the sub-target fault sample, and determine whether the calculated Euclidean distance is less than the domain radius of the current cycle; if so, determine that the fault sample is within the circular area, if not, determine that the fault sample is not within the circular area; until each fault sample is traversed, the number of fault samples in the circular area is obtained; determine whether the number of fault samples in the circular area is greater than the minimum number of neighbors of the current cycle; if so, determine that the sub-target fault sample is a core object, and determine all density-connected fault samples starting from the core object;

[0094] For each determined core object, the following steps are performed: traversing all core objects within the domain radius of the current loop of the core object, and finding all fault samples that are densely connected to these core objects to form a corresponding cluster, and determining the fault mode of the cluster based on the fault label of each fault sample contained in the cluster;

[0095] For each determined cluster, the following steps are performed: traverse each fault sample in the cluster and determine whether the fault mode corresponding to the fault label of the traversed fault sample is consistent with the fault mode of the cluster; if so, retain the fault sample; if not, remove the fault sample from the cluster; until each fault sample in the cluster is traversed and the accurate number of samples in the cluster is obtained;

[0096] S3, based on the number of accurate samples in each cluster, calculate the fault recognition accuracy of the current cycle, and execute S4;

[0097] S4, judging whether the fault recognition accuracy of the current cycle is greater than the accuracy corresponding to the current best position of the current particle, if so, updating the best position of the current particle; judging whether the fault recognition accuracy of the current cycle is greater than the accuracy corresponding to the current cycle best position of the current particle group, if so, updating the best position of the current particle group;

[0098] S5, determine whether the preset stop condition is met; if not, update the position and speed of each particle, increase the number of iterations by 1, and execute S2 to S4 repeatedly until the preset stop condition is met to obtain the final best position of the particle swarm;

[0099] S6, taking the domain radius and the minimum neighbor number corresponding to the final best position of the particle swarm as the optimal domain radius and the optimal minimum neighbor number of the clustering algorithm.

[0100] In some implementations, the fault recognition rate is calculated using the following formula:

[0101]

[0102] Where c is the fault recognition rate; j = 1, 2...N, N is the number of fault modes; w j N is the hazard level of the jth failure mode; j is the total number of fault samples contained in the cluster corresponding to the jth fault mode; is the number of correct fault samples contained in the cluster corresponding to the jth fault mode.

[0103] In some embodiments, the preset stop condition is:

[0104] The accuracy corresponding to the best position of the current cycle of the particle swarm is greater than the accuracy threshold;

[0105] And / or, the number of iterations is greater than a threshold number.

[0106] In some implementations, the computing unit 306 is configured to perform the following operations:

[0107] Traverse each cluster and execute the following command for each cluster:

[0108] Calculate the distance between the on-track data and each core object in the cluster; and determine whether there is a core object that satisfies the distance from the on-track data to the core object is less than the optimal domain radius; if so, determine that the on-track data belongs to the cluster; if not, traverse the next cluster until the cluster to which the track data belongs is determined.

[0109] It should be noted that the spacecraft control system fault identification device based on multi-source information fusion provided in the above embodiment is only illustrated by the division of the above functional modules. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the spacecraft control system fault identification device based on multi-source information fusion provided in the above embodiment and the spacecraft control system fault identification method embodiment based on multi-source information fusion belong to the same concept. The specific implementation process is detailed in the method embodiment and will not be repeated here.

[0110] The embodiment of the present application also provides a computer device, please refer to Figure 3 The computer device includes a processor and a memory, wherein the memory stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, at least one program, a code set or an instruction set is loaded and executed by the processor to implement the spacecraft control system fault identification method based on multi-source information fusion provided by the above-mentioned method embodiments.

[0111] An embodiment of the present application also provides a computer-readable storage medium, on which is stored at least one instruction, at least one program, code set or instruction set, and the at least one instruction, at least one program, code set or instruction set is loaded and executed by a processor to implement the spacecraft control system fault identification method based on multi-source information fusion provided in the above-mentioned method embodiments.

[0112] An embodiment of the present application also provides a computer program product, which includes a computer program. A processor of a computer device reads the computer program from a computer-readable storage medium, and the processor executes the computer program, so that the computer device executes the spacecraft control system fault identification method based on multi-source information fusion described in any of the above embodiments.

[0113] For the convenience of description, the above system or device is described by dividing it into various modules or units according to its functions. Of course, when implementing the present application, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0114] It can be known from the description of the above implementation methods that those skilled in the art can clearly understand that the present application can be implemented by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present application can be essentially or partly contributed to the prior art in the form of a software product, which can be stored in a storage medium such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in the various embodiments of the present application or certain parts of the embodiments.

[0115] Finally, it should be noted that, in this article, relational terms such as first, second, third and fourth are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the statement "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0116] The above is only a preferred implementation of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A method for fault identification of spacecraft control system based on multi-source information fusion, characterized in that: The method comprises: The data of the spacecraft during the development and testing process and the historical data on orbit are multi-source fused to obtain multiple fault samples; each of the fault samples is provided with a fault label, and each of the fault labels is used to characterize the fault mode corresponding to the fault sample; Performing normalization processing on each of the fault samples to obtain a normalized fault sample; Based on a preset clustering algorithm, the normalized fault samples are clustered to obtain multiple clusters and the fault mode of each cluster; wherein each cluster includes at least one core object; For the on-orbit data of the spacecraft at any moment, the distance between the on-orbit data and the core object in each cluster is calculated, so as to determine the cluster to which the on-orbit data belongs based on the calculated distance; The failure mode corresponding to the cluster to which the on-track data belongs is determined as the identification result of the on-track data.

2. The method according to claim 1, characterized in that: The preset clustering algorithm is the DBSCAN algorithm; The normalized fault samples are clustered based on a preset clustering algorithm to obtain multiple clusters and the fault mode of each cluster, including: The DBSCAN algorithm is optimized based on a particle swarm algorithm to obtain an optimal domain radius and an optimal minimum neighbor number of the DBSCAN algorithm; Taking each fault sample as a target fault sample in turn, and determining a circular area with the target fault sample as the center and the optimal domain radius as the radius; traversing other fault samples, for each fault sample traversed, calculating the Euclidean distance between the fault sample and the target fault sample, and judging whether the calculated Euclidean distance is less than the optimal domain radius; if so, judging that the fault sample is within the circular area, if not, judging that the fault sample is not within the circular area; until each fault sample is traversed, the number of fault samples within the circular area is obtained; Determine whether the number of fault samples in the circular area is greater than the optimal minimum neighbor number; if so, determine that the target fault sample is a core object, and determine all density-connected fault samples starting from the core object; For each determined core object, the following steps are performed: traverse all core objects within the optimal domain radius of the core object, and find all fault samples that are densely connected to these core objects to form a corresponding cluster, and determine the fault mode of the cluster based on the fault label of each fault sample contained in the cluster.

3. The method according to claim 2, characterized in that The determining the failure mode of the cluster based on the failure label of each failure sample included in the cluster includes: Count the number of fault samples corresponding to each fault label in the cluster respectively; The fault mode represented by the fault label with the largest number of corresponding fault samples is determined as the fault mode of the cluster.

4. The method according to claim 2, characterized in that: The particle swarm algorithm is used to optimize the preset clustering algorithm to obtain the optimal domain radius and the optimal minimum neighbor number of the clustering algorithm, including: S1, initializing the particle swarm parameters, which include the number of particles, the initial best position of the particle swarm, the initial position of each particle, the initial speed and the initial best position of the particle; S2, traverse each particle in the particle group in turn, and for each traversed particle, execute: take each fault sample as a sub-target fault sample in turn, and determine a circular area with the sub-target fault sample as the center and the domain radius of the current cycle as the radius; traverse other fault samples, and for each traversed fault sample, calculate the Euclidean distance between the fault sample and the sub-target fault sample, and determine whether the calculated Euclidean distance is less than the domain radius of the current cycle; if so, determine that the fault sample is within the circular area, if not, determine that the fault sample is not within the circular area; until each fault sample is traversed, the number of fault samples in the circular area is obtained; determine whether the number of fault samples in the circular area is greater than the minimum number of neighbors of the current cycle; if so, determine that the sub-target fault sample is a core object, and determine all density-connected fault samples starting from the core object; For each determined core object, the following steps are performed: traversing all core objects within the domain radius of the current loop of the core object, and finding all fault samples that are densely connected to these core objects to form a corresponding cluster, and determining the fault mode of the cluster based on the fault label of each fault sample contained in the cluster; For each determined cluster, the following steps are performed: traverse each fault sample in the cluster and determine whether the fault mode corresponding to the fault label of the traversed fault sample is consistent with the fault mode of the cluster; if so, retain the fault sample; if not, remove the fault sample from the cluster; until each fault sample in the cluster is traversed and the accurate number of samples in the cluster is obtained; S3, based on the number of accurate samples in each cluster, calculate the fault recognition accuracy of the current cycle, and execute S4; S4, judging whether the fault recognition accuracy of the current cycle is greater than the accuracy corresponding to the current best position of the current particle, if so, updating the best position of the current particle; judging whether the fault recognition accuracy of the current cycle is greater than the accuracy corresponding to the current cycle best position of the current particle group, if so, updating the best position of the current particle group; S5, determining whether the preset stop condition is reached; if not, updating the position and speed of each particle, increasing the number of iterations by 1, and looping through S2 to S4 until the preset stop condition is reached, thereby obtaining the final best position of the particle swarm; S6, taking the domain radius and the minimum neighbor number corresponding to the final best position of the particle swarm as the optimal domain radius and the optimal minimum neighbor number of the clustering algorithm.

5. The method according to claim 4, characterized in that The fault recognition rate is calculated using the following formula: Where c is the fault recognition rate; j = 1, 2...N, N is the number of fault modes; w j N is the hazard level of the jth failure mode; j is the total number of fault samples contained in the cluster corresponding to the jth fault mode; is the number of correct fault samples contained in the cluster corresponding to the jth fault mode.

6. The method according to claim 4, characterized in that The preset stop condition is: The accuracy rate corresponding to the best position of the current cycle of the particle swarm is greater than the accuracy rate threshold; And / or, the number of iterations is greater than a threshold number.

7. The method according to claim 1, characterized in that The step of calculating the distance between the on-orbit data and the core object in each cluster for the on-orbit data of the spacecraft at any time, so as to determine the cluster to which the on-orbit data belongs based on the calculated distance, includes: Traverse each cluster and execute the following command for each cluster: Calculate the distance between the on-track data and each core object in the cluster; and determine whether there is a core object that satisfies the distance from the on-track data to the core object is less than the optimal domain radius; if so, determine that the on-track data belongs to the cluster; if not, traverse the next cluster until the cluster to which the track data belongs is determined.

8. A spacecraft control system fault identification device based on multi-source information fusion, characterized in that: The device comprises: A fusion unit is used to fuse the data of the spacecraft during the development and testing process and the historical data on orbit to obtain multiple fault samples; each of the fault samples is provided with a fault label, and each of the fault labels is used to characterize the fault mode corresponding to the fault sample; A processing unit, used for performing normalization processing on each of the fault samples to obtain a normalized fault sample; A clustering unit, used for clustering the normalized fault samples based on a preset clustering algorithm to obtain multiple clusters and a fault mode of each cluster; wherein each cluster includes at least one core object; a calculation unit, configured to calculate, for the on-orbit data of the spacecraft at any moment, a distance between the on-orbit data and a core object in each cluster, so as to determine the cluster to which the on-orbit data belongs based on the calculated distance; The determination unit is used to determine the failure mode corresponding to the cluster to which the on-track data belongs as the identification result of the on-track data.

9. A computer device, characterized in that: The computer device includes a memory and a processor, the memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory to implement the steps of any one of the methods described in claims 1-7.

10. A computer-readable storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method described in any one of claims 1 to 7 are implemented.

Images