A method for diagnosing and handling faults of a drag-free satellite

Abstract

The present application relates to the technical field of spacecraft operation fault diagnosis, in particular to a fault diagnosis and disposal method for a drag-free satellite. The method comprises the following steps: obtaining real-time measurement state variables of a dynamic system of the drag-free satellite, performing fault diagnosis and integration on the measurement state variables to obtain an actual binary diagnosis vector, matching the actual binary diagnosis vector with fault information units stored in an expert knowledge base one by one, and feeding back all fault information units having a matching degree with the actual binary diagnosis vector, and calling the matched fault information units one by one to obtain and call the precision threshold diagnosis scheme and disposal scheme corresponding to the fault information units. The present application is based on the fault diagnosis and disposal of the drag-free satellite based on the modularized expert knowledge base, so that the constructed expert knowledge base can meet the requirements of the fault diagnosis of the auxiliary drag-free gravitational wave detection satellite.

Description

Technical Field

[0001] This invention relates to the field of spacecraft operation fault diagnosis technology, and in particular to a method for diagnosing and handling faults in a dragless satellite. Background Technology

[0002] The space-based gravitational wave detection system consists of a formation of three dragless satellites, each carrying two test masses, achieving nanometer-level relative motion. In addition to the basic subsystems maintaining the satellites' normal operation, the dragless satellite platform also carries gravitational wave detection payloads such as electrode cages and test masses. Therefore, the gravitational wave detection platform is a typical high-precision, high-complexity mechanical system, which makes it susceptible to various potential failure modes. Therefore, constructing an expert knowledge base that pre-stores known failure information of the gravitational wave detection platform can assist in faster fault matching and provide known feasible fault handling solutions to users for decision-making when failures occur.

[0003] Due to the complexity of fault types and the variety of fault handling solutions, expert systems have become one of the important technologies for assisting in the fault diagnosis of complex systems, and have been widely used in many technology-intensive fields such as aerospace, industrial electronics, and medicine. Expert systems use a "knowledge + reasoning" process to solve interdisciplinary problems that require complex knowledge to solve. Therefore, the expert knowledge base, which stores all the knowledge of the expert system, is the core; that is, the construction method and knowledge management system of the expert knowledge base directly affect the accuracy and efficiency of the expert system. The expert knowledge base is an electronic record of expert knowledge entities in the corresponding domain. Knowledge representation is the process of describing human knowledge entities using symbols that can be recognized and processed by computers. For expert systems, the knowledge representation method when constructing the knowledge base directly affects the convenience and query efficiency of the knowledge base. Therefore, the knowledge representation process is one of the core issues in constructing an expert knowledge base. Expert knowledge bases based on traditional knowledge representation methods such as generation systems, framework structures, and semantic networks are difficult to apply to the switching between various fault diagnosis modes of drag-free gravitational wave detection satellite platforms, and also cannot meet the needs for convenient retrieval and access to expert knowledge. Summary of the Invention

[0004] The technical problem to be solved by the embodiments of the present invention is to provide a method for diagnosing and handling faults in dragless satellites, so as to solve the problem that the expert knowledge base formed by traditional knowledge representation methods in the prior art cannot be applied to the fault diagnosis of dragless gravitational wave detection satellites.

[0005] This invention discloses a method for diagnosing and handling faults in dragless satellites, including:

[0006] Acquire known fault information units of the dragless satellite dynamics system, as well as sample binary diagnostic vectors, fine threshold diagnostic schemes, and handling schemes corresponding to the fault information units, and construct an expert knowledge base;

[0007] The real-time measurement state quantities of the dragless satellite dynamics system are obtained, and the measurement state quantities are used for fault diagnosis and integration to obtain the actual binary diagnostic vector.

[0008] The actual binary diagnostic vector is matched with the sample binary diagnostic vector to obtain all the fault information units that have a matching degree with the current actual binary diagnostic vector;

[0009] All the matched fault information units are called one by one, and the corresponding precision threshold diagnosis scheme and treatment scheme are obtained and called.

[0010] Optionally, the method for constructing the expert knowledge base includes:

[0011] A unique code is established for each of the fault information units;

[0012] The encoding of the fault information unit is associated with the sample binary diagnostic vector, the fine threshold diagnostic scheme, and the treatment scheme corresponding to the fault information unit.

[0013] Based on the set of associated codes, the fault information unit and the correspondence between it and the sample binary diagnostic vector, the precision threshold diagnostic scheme, and the treatment scheme are established and stored to form an expert knowledge base.

[0014] Optionally, the method of matching the actual binary diagnostic vector with the sample binary diagnostic vector to obtain all the fault information units that have a matching degree with the current actual binary diagnostic vector includes:

[0015] The actual binary diagnostic vector is matched one by one with the fault information units stored in the expert knowledge base, and the matching degree between the actual binary diagnostic vector and the sample binary diagnostic vector corresponding to the fault information unit is calculated.

[0016] Based on the calculated matching degree, a set of codes that have a matching degree with the current actual binary diagnostic vector is obtained, and the fault information unit corresponding to the code is obtained by calling the codes in the set one by one.

[0017] Optionally, the method for calculating the matching degree includes:

[0018] Obtain the sample binary coarse threshold diagnosis vector obtained by the fault information unit under coarse threshold diagnosis;

[0019] The matching degree between the actual binary diagnostic vector and the sample binary coarse threshold diagnostic vector is calculated using the matching degree calculation formula, which is:

[0020]

[0021] Where η is the matching degree, and d(t) is the actual binary diagnostic vector. (t) is the sample binary coarse threshold diagnosis vector obtained by the current fault information unit under coarse threshold diagnosis, T is the transpose of the table vector, and t is the current sampling time.

[0022] Optionally, the method for establishing and storing the correspondence between the fault information unit and the sample binary diagnostic vector, the precision threshold diagnostic scheme, and the treatment scheme includes:

[0023] The sample binary diagnostic vectors obtained by the fault information unit under coarse threshold and fine threshold diagnosis are obtained respectively;

[0024] Based on the sample binary diagnostic vector, precise threshold diagnostic scheme, and treatment scheme corresponding to the fault information unit, corresponding strings are generated respectively, and a storage partition for the fault information unit is established. The expression for the storage partition of the fault information unit is:

[0025]

[0026] c i ∈[c1,......,c f ]

[0027] in, For fault information unit, c i Here, f represents the encoding of the fault information unit, and f is the total number of fault information units stored in the expert knowledge base. A string containing the current fault information unit encoding. This is a string containing the sample binary coarse threshold diagnosis vector obtained by the current fault information unit under coarse threshold diagnosis. This is a string containing the sample binary fine threshold diagnostic vector obtained by the current fault information unit under fine threshold diagnosis. For a string containing a precision threshold diagnostic scheme, A string representing the disposal plan.

[0028] Optionally, the method of sequentially calling all the matched fault information units and obtaining and calling the precise threshold diagnosis scheme and treatment scheme corresponding to the fault information unit includes:

[0029] Based on the number of matched fault information units, the corresponding code of the fault information unit, as well as the precise threshold diagnosis scheme and handling scheme, are fed back:

[0030] When a unique fault information unit is matched, the code corresponding to the unique fault information unit is fed back, and the precise threshold diagnosis scheme and treatment scheme corresponding to the code are obtained by calling the code.

[0031] When multiple fault information units are matched, the codes corresponding to the multiple fault information units are fed back, and the precise threshold diagnosis scheme and treatment scheme corresponding to each code are obtained by calling the multiple codes one by one.

[0032] Optionally, the method for fault diagnosis and integration of the measured state quantities to obtain the actual binary diagnostic vector includes:

[0033] Obtain the measurement status set of multiple channels in the measurement status quantity under normal and abnormal conditions;

[0034] A coarse threshold is set, and the measurement state set of each channel data is mapped to a one-dimensional "0-1" representation to construct the actual binary diagnostic vector of the channel data;

[0035] When a measurement state in the measurement state set is normal data that does not exceed the coarse threshold, the measurement state is mapped to a "0" element.

[0036] When a measurement state in the measurement state set is fault data exceeding the coarse threshold, the measurement state is mapped to a "1" element.

[0037] Optionally, the method for obtaining known fault information units of the dragless satellite dynamics system, and the corresponding sample binary diagnostic vector, fine threshold diagnostic scheme, and handling scheme for the fault information units, includes:

[0038] Acquire the measurement status set of the fault information unit in normal and abnormal states;

[0039] A coarse threshold is set to diagnose the measurement state set of the fault information unit, and a sample binary coarse threshold diagnosis vector of the fault information unit is constructed.

[0040] Obtain the diagnostic results of the measurement state set of the fault information unit using a coarse threshold, set a fine threshold to diagnose the measurement state set of the fault information unit, and construct a sample binary fine threshold diagnostic vector for the fault information unit.

[0041] Obtain the diagnostic results of the measurement state set of the fault information unit using a precise threshold, and establish a precise threshold diagnostic scheme and treatment scheme for the fault information unit by analyzing and verifying the diagnostic results.

[0042] Optionally, it also includes a method for handling situations where the actual binary diagnostic vector does not match the fault information unit:

[0043] Feedback is provided on the actual binary diagnostic vector that does not match the fault information unit, an unknown fault information unit is established, and the actual binary diagnostic vector is associated with the sample binary coarse threshold diagnostic vector of the unknown fault information unit.

[0044] A precise threshold is set to diagnose the measurement state set of the fault information unit, a sample binary precise threshold diagnosis vector of the fault information unit is constructed, and the diagnosis results are analyzed and verified to establish a precise threshold diagnosis scheme and a handling scheme for the unknown fault information unit.

[0045] Establish the correspondence between the unknown fault information unit and the sample binary coarse threshold diagnosis vector, sample binary fine threshold diagnosis vector, fine threshold diagnosis scheme and treatment scheme, and store and update the expert knowledge base.

[0046] Optionally, the method for obtaining the measured state quantities of the dragless satellite dynamics system includes:

[0047] The measured state quantities of the drag-free satellite dynamics system are acquired in real time at a fixed sampling frequency. These measured state quantities include:

[0048] The linear motion position vector and velocity vector of the gravitational wave detection platform;

[0049] Angular motion position vector and angular velocity vector of the gravitational wave detection platform;

[0050] The first test mass line motion position vector and velocity vector, as well as its angular motion position vector and velocity vector, were located within the gravitational wave detection platform.

[0051] The position and velocity vectors of the second test mass line motion within the gravitational wave detection platform, as well as the position and velocity vectors of its angular motion;

[0052] The rotational angular displacement and rotational angular velocity of the electrode cage inside the gravitational wave detection platform.

[0053] Compared with the prior art, the advantages of the dragless satellite fault diagnosis and handling method provided in this invention are as follows:

[0054] The system acquires real-time measured state quantities of the dragless satellite dynamics system, performs fault diagnosis and integration on these measured state quantities to obtain an actual binary diagnostic vector, matches each actual binary diagnostic vector with the fault information units stored in the constructed expert knowledge base, and feeds back all fault information units that match the current actual binary diagnostic vector. By calling each matched fault information unit, the system obtains and calls the corresponding fine threshold diagnostic and treatment schemes for the fault information units, thereby completing the dragless satellite fault diagnosis and treatment based on the modular expert knowledge base. This enables the constructed expert knowledge base to meet the fault diagnosis needs of dragless gravitational wave detection satellites. Attached Figure Description

[0055] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. In the accompanying drawings:

[0056] Figure 1 A schematic block diagram illustrating the steps of the dragless satellite fault diagnosis and handling method provided in this embodiment of the invention;

[0057] Figure 2 A schematic flowchart of the system flow diagram for the method for diagnosing and handling faults in a dragless satellite provided in an embodiment of the present invention;

[0058] Figure 3 This is a schematic diagram of the coarse threshold fault diagnosis matching result of the current actual binary diagnostic vector provided in the embodiment of the present invention;

[0059] Figure 4 A schematic diagram illustrating the results of the feedback precision threshold diagnostic scheme provided in an embodiment of the present invention;

[0060] Figure 5 This is a schematic diagram illustrating the results of the feedback-based fine threshold handling scheme provided in an embodiment of the present invention. Detailed Implementation

[0061] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. The preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0062] like Figure 1 and Figure 2 As shown, this invention discloses a method for diagnosing and handling faults in dragless satellites, comprising:

[0063] S1. Obtain known fault information units of the dragless satellite dynamics system, as well as sample binary diagnostic vectors, fine threshold diagnostic schemes and handling schemes corresponding to the fault information units, and construct an expert knowledge base;

[0064] S2. Obtain the real-time measurement state quantities of the dragless satellite dynamics system, perform fault diagnosis and integration on the measurement state quantities to obtain the actual binary diagnostic vector;

[0065] S3. Match the actual binary diagnostic vector with the sample binary diagnostic vector to obtain all fault information units that have a matching degree with the current actual binary diagnostic vector;

[0066] S4. Call all matched fault information units one by one, and obtain and call the corresponding precision threshold diagnosis scheme and handling scheme for each fault information unit.

[0067] By implementing the above method, real-time measured state quantities of the dragless satellite dynamics system are obtained. Fault diagnosis and integration of these measured state quantities yield an actual binary diagnostic vector. This actual binary diagnostic vector is then matched one by one with the fault information units stored in the constructed expert knowledge base. All fault information units with a matching degree to the current actual binary diagnostic vector are fed back. By calling each matched fault information unit, the corresponding precise threshold diagnostic and handling schemes are obtained and invoked. This completes the dragless satellite fault diagnosis and handling based on the modular expert knowledge base, enabling the constructed expert knowledge base to meet the fault diagnosis requirements of dragless gravitational wave detection satellites. The method of this embodiment can be applied to dragless satellite fault matching and handling based on a modular expert knowledge base.

[0068] Furthermore, methods for constructing expert knowledge bases include:

[0069] A unique code is established for each fault information unit;

[0070] The encoding of the fault information unit is associated with the corresponding sample binary diagnostic vector, fine threshold diagnostic scheme and treatment scheme.

[0071] Based on the set of associated codes, fault information units are established and stored, along with the correspondence between sample binary diagnostic vectors, precision threshold diagnostic schemes, and treatment schemes to form an expert knowledge base.

[0072] Furthermore, the method for matching the actual binary diagnostic vector with the sample binary diagnostic vector to obtain all fault information units that have a matching degree with the current actual binary diagnostic vector includes:

[0073] The actual binary diagnostic vector is matched one by one with the fault information units stored in the expert knowledge base, and the matching degree between the actual binary diagnostic vector and the corresponding sample binary diagnostic vector of the fault information unit is calculated.

[0074] Based on the calculated matching degree, a set of codes that match the feedback and the current actual binary diagnostic vector is obtained, and the fault information unit corresponding to the code is obtained by calling the codes in the set one by one.

[0075] The fault information units stored in the expert knowledge base can be represented as {Sc1,......,Sci}, and the encoding set of the stored fault information units can be represented as: and Let q ∈ N be the total number of fault information units stored in the expert knowledge base. + This represents the number of all fault information units that have a matching degree with the current actual binary diagnostic vector.

[0076] Furthermore, the methods for calculating the matching degree include:

[0077] The fault information acquisition unit obtains a sample binary coarse threshold diagnosis vector under coarse threshold diagnosis.

[0078] The matching degree between the actual binary diagnostic vector and the sample binary coarse threshold diagnostic vector is calculated using the matching degree calculation formula:

[0079]

[0080] Where η is the matching degree, and d(t)∈N38×1 is the actual binary diagnostic vector. Let T be the sample binary coarse threshold diagnosis vector obtained by the current fault information unit under coarse threshold diagnosis, where T is the transpose of the table vector and t is the current sampling time.

[0081] Furthermore, the method for establishing and storing the correspondence between fault information units and sample binary diagnostic vectors, precise threshold diagnostic schemes, and treatment schemes includes:

[0082] The sample binary diagnostic vectors obtained by the fault information unit under coarse threshold and fine threshold diagnosis are acquired respectively;

[0083] Based on the sample binary diagnostic vector, precise threshold diagnostic scheme, and treatment scheme corresponding to the fault information unit, generate the corresponding strings, and establish the storage partition for the fault information unit. The expression for the storage partition of the fault information unit is:

[0084]

[0085]

[0086] in, For fault information unit, c i Here, f represents the encoding of the fault information unit, and f is the total number of fault information units stored in the expert knowledge base. A string containing the current fault information unit encoding. This is a string containing the sample binary coarse threshold diagnosis vector obtained by the current fault information unit under coarse threshold diagnosis. This is a string containing the sample binary fine threshold diagnostic vector obtained by the current fault information unit under fine threshold diagnosis. For a string containing a precision threshold diagnostic scheme, A string representing the disposal plan.

[0087] As described above, by returning the encoded set C = {c1, ..., c2} of fault information units that have a matching degree with the current actual binary diagnostic vector d(t), q}, call the stored fault information unit corresponding to each element in set C one by one. i∈{1,......q}, and the fault information unit The corresponding number c i Precise threshold diagnosis scheme for faults and troubleshooting plan Return to the user interface to complete the drag-free satellite fault matching and handling based on the modular expert knowledge base.

[0088] Furthermore, the method of sequentially calling all matched fault information units and obtaining and calling the corresponding precise threshold diagnostic and treatment schemes for each fault information unit includes:

[0089] Based on the number of matched fault information units, the corresponding codes for the fault information units, as well as the precise threshold diagnostic and handling schemes, are fed back:

[0090] When a unique fault information unit is matched, the code corresponding to the unique fault information unit is fed back, and the precise threshold diagnosis scheme and treatment scheme corresponding to the code are obtained by calling the code.

[0091] When multiple fault information units are matched, the codes corresponding to the multiple fault information units are fed back, and the precise threshold diagnosis scheme and treatment scheme corresponding to each code are obtained by calling the multiple codes one by one.

[0092] Furthermore, methods for fault diagnosis and integration of measured state quantities to obtain the actual binary diagnostic vector include:

[0093] Obtain the measured state quantity x(t)∈R 38×1 The measurement status set of multiple channels in normal and abnormal states;

[0094] Set a coarse threshold and map the measurement state set of each channel data to a one-dimensional "0-1" representation to construct the actual binary diagnostic vector of the channel data;

[0095] When a certain measurement state x in the measurement state set k ∈R(k∈[1,38]∩N, where N is the set of positive integers) represents the value that does not exceed the coarse threshold ts kWhen the data is normal (∈R), the measurement state is mapped to a "0" element representation;

[0096] When a certain measurement state x in the measurement state set k ∈R represents values ​​exceeding the coarse threshold ts k When fault data ∈R is used, the measurement status is mapped to a "1" element representation.

[0097] Furthermore, the methods for obtaining known fault information units of a dragless satellite dynamics system, as well as the corresponding sample binary diagnostic vectors, fine threshold diagnostic schemes, and handling schemes for these fault information units, include:

[0098] The measurement status set of the fault information unit in normal and abnormal states;

[0099] A coarse threshold is set to diagnose the measurement state set of the fault information unit, and a sample binary coarse threshold diagnosis vector of the fault information unit is constructed.

[0100] Obtain the diagnostic results of the measurement state set of the fault information unit using a coarse threshold, set a fine threshold to diagnose the measurement state set of the fault information unit, and construct a sample binary fine threshold diagnostic vector for the fault information unit.

[0101] The diagnostic results of the fault information unit measurement state set are obtained by using precise thresholds. The precise threshold diagnostic scheme and treatment scheme of the fault information unit are established by analyzing and verifying the diagnostic results.

[0102] Furthermore, it also includes a method for handling situations where the actual binary diagnostic vector does not match a fault information unit:

[0103] Feedback on the actual binary diagnostic vector of the fault information unit that did not match, establish the unknown fault information unit, and associate the actual binary diagnostic vector with the sample binary coarse threshold diagnostic vector of the unknown fault information unit;

[0104] A precise threshold is set to diagnose the measurement state set of the fault information unit, a sample binary precise threshold diagnosis vector of the fault information unit is constructed, and the diagnosis results are analyzed and verified to establish a precise threshold diagnosis scheme and handling scheme for unknown fault information units.

[0105] Establish the correspondence between the unknown fault information unit and the sample binary coarse threshold diagnosis vector, sample binary fine threshold diagnosis vector, fine threshold diagnosis scheme and treatment scheme, store and update the expert knowledge base.

[0106] Furthermore, methods for obtaining the measured state quantities of a dragless satellite dynamic system include:

[0107] The measured state quantities of the drag-free satellite dynamics system are acquired in real time at a fixed sampling frequency. The measured state quantities x(t)∈R38×1 include:

[0108] The position vector rB ∈ R3×1 and velocity vector of the gravitational wave detection platform

[0109] The angular motion position vector θB∈R3×1 and the angular velocity vector of the gravitational wave detection platform

[0110] The first test mass line motion position vector rtm1∈R3×1 and velocity vector within the gravitational wave detection platform and its angular motion position vector θtm1∈R3×1 and velocity vector

[0111] The position vector rtm2∈R3×1 and velocity vector of the second test mass line motion within the gravitational wave detection platform And its position vector θtm2∈R3×1 and velocity vector

[0112] The rotational angular displacement α∈R and rotational angular velocity of the electrode cage C inside the gravitational wave detection platform

[0113] Combination Figures 3-5 As shown, a preferred simulation implementation method is as follows:

[0114] The true one-dimensional "0-1" binary diagnostic vector used for testing is given by a random integer sequence.

[0115] In this embodiment, the expert knowledge base used for testing is configured to store f=11 groups of fault information units, namely {Sc1,......,Sc11}.

[0116] In this embodiment, taking a fault information unit coded as c1=1 as an example, the information stored in its fault information unit storage partition is set as follows:

[0117]

[0118]

[0119] Shut down the current body measurement system, enable the backup measurement system, switch to the fine threshold diagnostic mode, apply body test stimuli, and re-execute the fault diagnosis vector.

[0120] Shut down the faulty thruster and switch to fault-tolerant control mode.

[0121] See Figures 3-5, respectively, represent the matching result between the current fault and the stored fault information unit when the dragless gravitational wave detection satellite and the fault information unit stored in the expert knowledge base have a matching degree, including all possible fault types, as well as the precise threshold diagnosis scheme for each type of fault and the recommended handling scheme after the fault is diagnosed.

[0122] The dragless satellite fault diagnosis and handling method of this invention constructs an expert knowledge base with a generalized structure storing fault information units, based on generalized descriptive elements of dragless satellite fault information. This expert knowledge base is suitable for space gravitational wave detection missions and has a universal structure. This allows the constructed expert knowledge base to switch between different diagnostic modes with coarse and fine thresholds, and to return fault handling methods to the user. Furthermore, simulation examples are used to verify the fault matching effect and fault handling scheme recommendation capability of the constructed expert knowledge base.

Claims

1. A method for diagnosing and handling faults in a dragless satellite, characterized in that, The method for diagnosing and handling faults in dragless satellites includes: Acquire known fault information units of the dragless satellite dynamics system, as well as sample binary diagnostic vectors, precise threshold diagnostic schemes, and handling schemes corresponding to the fault information units, and construct an expert knowledge base; The real-time measurement state quantities of the dragless satellite dynamics system are obtained, and the measurement state quantities are used for fault diagnosis and integration to obtain the actual binary diagnostic vector. The actual binary diagnostic vector is matched with the sample binary diagnostic vector to obtain all the fault information units that have a matching degree with the current actual binary diagnostic vector; Call all the matched fault information units one by one, and obtain and call the precise threshold diagnosis scheme and treatment scheme corresponding to the fault information unit; The method of matching the actual binary diagnostic vector with the sample binary diagnostic vector to obtain all fault information units that have a matching degree with the current actual binary diagnostic vector includes: The actual binary diagnostic vector is matched one by one with the fault information units stored in the expert knowledge base, and the matching degree between the actual binary diagnostic vector and the sample binary diagnostic vector corresponding to the fault information unit is calculated. Based on the calculated matching degree, a set of codes that have matching degree with the actual binary diagnostic vector is obtained, and the fault information unit corresponding to the code is obtained by calling the codes in the set one by one. The method for calculating the matching degree includes: Obtain the sample binary coarse threshold diagnosis vector obtained by the fault information unit under coarse threshold diagnosis; The matching degree between the actual binary diagnostic vector and the sample binary coarse threshold diagnostic vector is calculated using the matching degree calculation formula, which is: in, For matching degree, This is the actual binary diagnostic vector. Let T be the sample binary coarse threshold diagnosis vector obtained by the current fault information unit under coarse threshold diagnosis, where T is the vector transpose and t is the current sampling time.

2. The method for diagnosing and handling faults in a dragless satellite according to claim 1, characterized in that, The methods for constructing the expert knowledge base include: A unique code is established for each of the fault information units; The encoding of the fault information unit is associated with the sample binary diagnostic vector, the precise threshold diagnostic scheme, and the treatment scheme corresponding to the fault information unit. Based on the set of associated codes, the fault information unit and the correspondence between it and the sample binary diagnostic vector, the precise threshold diagnostic scheme and the treatment scheme are established and stored to form an expert knowledge base.

3. The method for diagnosing and handling faults in a dragless satellite according to claim 2, characterized in that, The method for establishing and storing the correspondence between the fault information unit and the sample binary diagnostic vector, the precise threshold diagnostic scheme, and the treatment scheme includes: Obtain the sample binary diagnostic vectors obtained by the fault information unit under coarse threshold and fine threshold diagnosis respectively; Based on the sample binary diagnostic vector, precise threshold diagnostic scheme, and treatment scheme corresponding to the fault information unit, corresponding strings are generated respectively, and a storage partition for the fault information unit is established. The expression for the storage partition of the fault information unit is: in, For fault information unit, Encoding for fault information units, This refers to the total number of fault information units stored in the expert knowledge base. A string containing the current fault information unit encoding. This is a string containing the sample binary coarse threshold diagnosis vector obtained by the current fault information unit under coarse threshold diagnosis. This is a string containing the sample binary precise threshold diagnosis vector obtained by the current fault information unit under precise threshold diagnosis. A string containing a precise threshold diagnostic scheme. A string representing the disposal plan.

4. The method for diagnosing and handling faults in a dragless satellite according to claim 2, characterized in that, The method of sequentially calling all the matched fault information units and obtaining and calling the precise threshold diagnosis scheme and treatment scheme corresponding to the fault information unit includes: Based on the number of matched fault information units, the corresponding code of the fault information unit, as well as the precise threshold diagnosis scheme and handling scheme, are fed back: When a unique fault information unit is matched, the code corresponding to the unique fault information unit is fed back, and the precise threshold diagnosis scheme and treatment scheme corresponding to the code are obtained by calling the code. When multiple fault information units are matched, the codes corresponding to the multiple fault information units are fed back, and the precise threshold diagnosis scheme and treatment scheme corresponding to each code are obtained by calling the multiple codes one by one.

5. The method for diagnosing and handling faults in a dragless satellite according to claim 1, characterized in that, The method for fault diagnosis and integration of the measured state quantities to obtain the actual binary diagnostic vector includes: Obtain the measurement status set of multiple channels in the measurement status quantity under normal and abnormal conditions; A coarse threshold is set, and the measurement state set of each channel data is mapped to a one-dimensional "0-1" representation to construct the actual binary diagnostic vector of the channel data; When a measurement state in the measurement state set is normal data that does not exceed the coarse threshold, the measurement state is mapped to a "0" element. When a measurement state in the measurement state set is fault data exceeding the coarse threshold, the measurement state is mapped to a "1" element.

6. The method for diagnosing and handling faults in a dragless satellite according to claim 1, characterized in that, The method for acquiring known fault information units of a dragless satellite dynamics system, as well as the corresponding sample binary diagnostic vector, precise threshold diagnostic scheme, and handling scheme for the fault information units, includes: Acquire the measurement status set of the fault information unit in normal and abnormal states; A coarse threshold is set to diagnose the measurement state set of the fault information unit, and a sample binary coarse threshold diagnosis vector of the fault information unit is constructed. Obtain the diagnostic results of the measurement state set of the fault information unit using a coarse threshold, set a fine threshold to diagnose the measurement state set of the fault information unit, and construct a sample binary fine threshold diagnostic vector for the fault information unit. Obtain the diagnostic results of the measurement state set of the fault information unit using a precise threshold, and establish a precise threshold diagnostic scheme and treatment scheme for the fault information unit by analyzing and verifying the diagnostic results.

7. The method for diagnosing and handling faults in a dragless satellite according to claim 1, characterized in that, It also includes a method for handling situations where the actual binary diagnostic vector does not match the fault information unit: Feedback is provided on the actual binary diagnostic vector that does not match the fault information unit, an unknown fault information unit is established, and the actual binary diagnostic vector is associated with the sample binary coarse threshold diagnostic vector of the unknown fault information unit. A precise threshold is set to diagnose the measurement state set of the fault information unit, a sample binary precise threshold diagnosis vector of the fault information unit is constructed, and the diagnosis results are analyzed and verified to establish a precise threshold diagnosis scheme and a handling scheme for the unknown fault information unit. Establish the correspondence between the unknown fault information unit and the sample binary coarse threshold diagnosis vector, sample binary fine threshold diagnosis vector, fine threshold diagnosis scheme and treatment scheme, and store and update the expert knowledge base.

8. The method for diagnosing and handling faults in a dragless satellite according to claim 1, characterized in that, The method for obtaining the measured state quantities of a drag-free satellite dynamic system includes: The measured state quantities of the drag-free satellite dynamics system are acquired in real time at a fixed sampling frequency. These measured state quantities include: The linear motion position vector and velocity vector of the gravitational wave detection platform; The angular motion position vector and angular velocity vector of the gravitational wave detection platform; The first test mass line motion position vector and velocity vector, as well as its angular motion position vector and velocity vector, were located within the gravitational wave detection platform. The position and velocity vectors of the second test mass line motion within the gravitational wave detection platform, as well as the position and velocity vectors of its angular motion; The rotational angular displacement and rotational angular velocity of the electrode cage inside the gravitational wave detection platform.

Images