A Rule Knowledge Fault Diagnosis Method and System for Ground Measurement and Control Equipment

Abstract

The present invention discloses a method and system for fault diagnosis of regular knowledge of ground measurement and control equipment. The method includes obtaining fault phenomena, matching initial rules, initializing an event scheduling center, the event scheduling center publishing knowledge reasoning events, obtaining the configuration of rule knowledge and diagnostic data, executing the rule knowledge reasoning events, performing subsequent operations according to the inferred actions until the entire rule knowledge system finishes reasoning, and generating a diagnostic conclusion. By designing regular knowledge for fault diagnosis of ground measurement and control equipment, the present invention realizes dynamically configurable rule knowledge diagnosis. The update operation of diagnostic knowledge is simple and convenient, and the diagnostic method can be iteratively upgraded during actual use, solving the technical problems that the current research on fault diagnosis methods of ground measurement and control equipment is insufficient, cannot adapt to the increasingly tense form of ground measurement and control resources, and cannot meet the urgent PHM needs of users.

Description

Technical Field

[0001] The present invention relates to the technical field of fault diagnosis of aerospace ground measurement and control equipment, and particularly to a rule-based knowledge fault diagnosis method and system for ground measurement and control equipment. Background Art

[0002] In recent years, with the strengthening of space awareness, the number of on-orbit and planned-to-be-launched spacecraft has shown a blowout growth. Accordingly, the corresponding ground measurement and control requirements have increased rapidly. To meet the huge measurement and control requirements, it is necessary to continuously expand and construct ground measurement and control equipment, which brings huge challenges to the operation and maintenance of the equipment. Therefore, it has become extremely urgent to establish an automated and intelligent central-level Prognostic and Health Management (PHM) system. All along, the research and production of ground measurement and control equipment have been carried out in the form of scientific research projects, resulting in customized equipment models, without forming standardization, systematization, and scale, making the research and application development of health management in related fields slow.

[0003] Fault diagnosis is an important part of the PHM system. Its main purposes are to locate the underlying fault modes, analyze the fault mechanisms, provide disposal suggestions, etc., and provide key inputs for functions such as false alarm elimination, health assessment, fault prediction, and maintenance support. Common fault diagnosis methods include expert knowledge systems, machine learning, deep learning, etc., which have all been successfully applied in PHM systems in different fields. However, the application of these methods in the fault diagnosis of ground measurement and control equipment has not been fully explored: on the one hand, due to the lack of a data acquisition, transmission, storage, and processing platform, machine learning and deep learning methods cannot be used to train and use fault diagnosis models; on the other hand, since the equipment is customized in the form of projects, there are no conditions for accumulating expert knowledge from historical data.

[0004] In summary, the current research on fault diagnosis methods for ground measurement and control equipment is insufficient, unable to adapt to the increasingly tense form of ground measurement and control resources, nor to meet the urgent PHM needs of users. For this reason, the present invention proposes a rule-based knowledge fault diagnosis method for ground measurement and control equipment. Summary of the Invention

[0005] The main purpose of the present invention is to provide a rule-based knowledge fault diagnosis method and system for ground measurement and control equipment, aiming to solve the technical problems that the current research on fault diagnosis methods for ground measurement and control equipment is insufficient, unable to adapt to the increasingly tense form of ground measurement and control resources, nor to meet the urgent PHM needs of users.

[0006] To achieve the above object, the present invention provides a rule-based knowledge fault diagnosis method for ground measurement and control equipment, and the method includes the following steps:

[0007] S1: Obtain the fault phenomenon; wherein, the fault phenomenon includes the equipment code, equipment node, and fault description;

[0008] S2: According to the fault phenomenon, match the corresponding initial rule knowledge;

[0009] S3: Initialize the event scheduling center, initialize the rule knowledge queue to be inferred as the obtained initial rule knowledge, and initialize the execution event count to 0;

[0010] S4: Obtain the rule knowledge at the head of the rule knowledge queue to be inferred;

[0011] S5: Determine whether the rule knowledge exists. If so, execute step S6; if not, when the execution event count is 0, execute step S4, and when the execution event count is not 0, output the diagnostic conclusion;

[0012] S6: Obtain the configuration and diagnostic data corresponding to the rule knowledge;

[0013] S7: Put the configuration and diagnostic data into the event, publish the inference event, increment the execution event count by one, and return to execute step S4;

[0014] S8: When publishing the inference event, asynchronously execute the rule inference to obtain the inference conclusion, and after the inference is completed, decrement the execution event count by one; wherein, the inference conclusion includes the fault mode and actions;

[0015] S9: If the action is normal end or abnormal end, end the process branch; if the action is rule knowledge, execute step S10;

[0016] S10: Put the rule knowledge into the rule knowledge queue to be inferred in the event scheduling center, and execute step S4.

[0017] Optionally, the rule knowledge includes tests, rule algorithms, fault modes, and action sets.

[0018] Optionally, the tests include feature extraction algorithms and test algorithms.

[0019] Optionally, the feature extraction algorithm is used to map the time series data into real numbers and transmit them as inputs to the test algorithm.

[0020] Optionally, the test algorithm is used to map the multi-dimensional feature values into boolean values.

[0021] Optionally, the feature extraction algorithm and the test algorithm are saved in the database table in the form of drools rules strings.

[0022] Optionally, the rule algorithm is a mapping from the test results to the fault mode and action set, including the fault mode judgment logic and the action selection logic.

[0023] Optionally, perform rule reasoning, which specifically includes the following steps:

[0024] S11: Input the time series data and the length of the time segment;

[0025] S12: According to the length of the time slice, sequentially split the time series data into data segments. If the remaining data at the end is less than one segment, it is regarded as one segment;

[0026] S13: Traverse the data segments;

[0027] S14: Traverse the tests;

[0028] S15: According to the monitoring points of the test, extract the data corresponding to the test monitoring points from the data segments;

[0029] S16: Call the feature extraction algorithm to map the time series data into real numbers;

[0030] S17: Execute the test algorithm, input the features of each monitoring point into the test algorithm, and obtain the test results;

[0031] S18: Have all the tests been executed? If yes, execute step S18; if no, jump to S14.

[0032] S19: Execute the fault mode judgment logic, substitute the test results into the fault mode judgment logic, and if the fault mode occurs, output the fault mode;

[0033] S20: Execute the action selection logic, substitute the test results into the action selection logic, and output the actions to continue execution;

[0034] S21: Judge whether a fault has occurred. If yes, execute step S23; if no, execute step S22;

[0035] S22: Judge whether the traversal of the data segments is over. If yes, execute step S23; if no, jump to S13.

[0036] S23: Output the fault mode and actions.

[0037] In addition, to achieve the above object, the present invention also provides a rule - knowledge fault diagnosis system for ground measurement and control equipment, which is applied to the rule - knowledge fault diagnosis method for ground measurement and control equipment as described above. The system includes:

[0038] An algorithm management module, which is used for the addition, deletion, query, modification and testing of the feature extraction algorithm and the test algorithm;

[0039] A data interaction module, which is used to obtain diagnostic data, test configurations and rule - knowledge configurations;

[0040] An algorithm engine module for setting algorithm inputs, executing algorithm logic, and obtaining algorithm results;

[0041] A knowledge inference module for parsing rule knowledge configurations, executing tests, executing rule algorithms, and generating inference conclusions;

[0042] A diagnostic process module for obtaining diagnostic phenomena, matching initial rules, obtaining rule knowledge configurations and diagnostic data, executing rule knowledge inference, scheduling inference events, and confirming diagnostic conclusions.

[0043] A method and system for diagnosing rule knowledge faults of ground measurement and control equipment proposed in an embodiment of the present invention. The method includes obtaining fault phenomena, matching initial rules, initializing an event scheduling center, the event scheduling center publishing knowledge inference events, obtaining the configuration and diagnostic data of rule knowledge, executing rule knowledge inference events, performing subsequent operations according to the inferred actions until the entire rule knowledge system inference ends, and generating diagnostic conclusions. By designing rule knowledge for diagnosing faults in ground measurement and control equipment, the present invention realizes dynamically configurable rule knowledge diagnosis. The update operation of diagnostic knowledge is simple and convenient, and the diagnostic method can be iteratively upgraded during actual use, solving the technical problems that the current research on fault diagnosis methods for ground measurement and control equipment is insufficient, cannot adapt to the increasingly tense form of ground measurement and control resources, and cannot meet the urgent PHM needs of users. Description of the Drawings

[0044] Figure 1 Schematic diagram of the process of the method for diagnosing rule knowledge of ground measurement and control equipment;

[0045] Figure 2 Schematic diagram of the inference process of diagnostic rule knowledge for ground measurement and control equipment;

[0046] Figure 3 Schematic diagram of the structure of diagnostic rule knowledge for ground measurement and control equipment;

[0047] Figure 4 Schematic diagram of the diagnostic rule knowledge system for ground measurement and control equipment;

[0048] Figure 5 Schematic diagram of the functional modules of the diagnostic rule knowledge system for ground measurement and control equipment;

[0049] Figure 6 Schematic diagram of the algorithm engine of the method for diagnosing rule knowledge of ground measurement and control equipment.

[0050] The realization, functional features, and advantages of the object of the present invention will be further described with reference to the embodiments and the accompanying drawings. Detailed Embodiments

[0051] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0052] At present, in the related technical fields, the research on the existing fault diagnosis methods for ground measurement and control equipment is insufficient, which cannot adapt to the increasingly tense form of ground measurement and control resources, nor can it meet the urgent PHM needs of users.

[0053] To solve this problem, various embodiments of the rule-based knowledge fault diagnosis method and system for ground measurement and control equipment of the present invention are proposed. The rule-based knowledge fault diagnosis method and system for ground measurement and control equipment provided by the present invention realize dynamic and configurable rule-based knowledge diagnosis by designing rule-based knowledge for the fault diagnosis of ground measurement and control equipment. The diagnosis knowledge update operation is simple, and the diagnosis method can be iteratively upgraded during actual use, solving the technical problems that the existing research on the fault diagnosis method for ground measurement and control equipment is insufficient, cannot adapt to the increasingly tense form of ground measurement and control resources, and cannot meet the urgent PHM needs of users.

[0054] An embodiment of the present invention provides a rule-based knowledge fault diagnosis method for ground measurement and control equipment, including the following steps:

[0055] 1. Obtain the fault phenomenon. The fault phenomenon describes the surface phenomenon of the fault, including fields such as equipment code, equipment node, and fault description.

[0056] 2. Match the initial rules. Since the rule knowledge action set can configure the rule knowledge, all rule knowledge can be regarded as a networked rule knowledge system. The matching conditions are that the equipment code and equipment node are the same and the fault description of the fault phenomenon is similar to the fault mode fault description of the rule knowledge. The matched rule knowledge is the starting point for the inference of the rule knowledge system. The present invention allows the inference process to execute in parallel from multiple starting points.

[0057] 3. Initialize the event scheduling center. The event scheduling center is mainly responsible for the inference scheduling of the rule knowledge system, that is, generating and publishing rule knowledge inference events and counting the events in execution. The event scheduling center contains a queue of rule knowledge to be inferred, which is initialized to the matched rule knowledge in this step. The count of events in execution is initialized to 0.

[0058] 4. The event scheduling center publishes knowledge inference events. The scheduling strategy is to sequentially publish the rule knowledge in the queue to be inferred. Before publishing, it is necessary to prepare the rule knowledge configuration and diagnostic data for the inference event.

[0059] 5. Obtain the configuration and diagnostic data of the rule knowledge. When preparing the rule knowledge inference event to be published, read the configuration of the rule knowledge and the required inference data from the database. The rule knowledge configuration is obtained through the rule number, and the inference data is obtained by querying according to the start / event of the diagnosis and the input of the test of the rule knowledge.

[0060] 6. Execute the rule knowledge inference event. First, execute the feature extraction and test algorithms for the test, then execute the fault mode judgment logic of the rule algorithm, and finally execute the action selection logic.

[0061] 7. Perform subsequent operations based on the inferred actions. If it is "end", output the fault mode to the diagnostic conclusion and normally end the current branch; if it is "uncertain", passively end the current branch; if it is "rule knowledge", put it into the inference queue of the scheduling center.

[0062] 8. Repeat steps 4 to 7 until the entire rule knowledge system inference ends.

[0063] 9. Generate the diagnostic conclusion. Take all the fault modes output during the inference process as the diagnostic conclusion.

[0064] It should be noted that this embodiment provides a rule knowledge diagnosis method for ground TT&C equipment, which is applicable to configuration-based automated fault location.

[0065] Among them, the rule knowledge for ground TT&C equipment fault diagnosis includes tests, rule algorithms, fault modes, and action sets.

[0066] Rule knowledge consists of four parts: tests, rule algorithms, fault modes, and action sets. Rule knowledge contains several tests, which should have clear physical meanings, good interpretability, and represent a condition for the occurrence of a fault mode; the rule algorithm is the mapping from test results to fault modes and action sets, including fault mode judgment logic and action selection logic; at most one fault mode exists in one piece of rule knowledge; the action set defines the scope of the next operation, including (several) rule knowledge, end, and uncertain.

[0067] The rule knowledge test consists of an input, a feature extraction algorithm, and a rule algorithm. The input is the time series data of the monitoring point, reported by the equipment end; the feature extraction maps the time series data of the monitoring point into real numbers as the input of the test algorithm. The test algorithm maps the multi-dimensional feature values into a boolean value, that is, the test result, indicating whether the test passes or fails.

[0068] The rule algorithm is the mapping from test results to fault modes and action sets, supporting basic logical operations such as AND, OR, NOT, XOR, and voting. The rule algorithm is divided into fault mode judgment logic and action selection logic. The fault mode judgment logic is mainly the mixed logical operation on the test results. When the operation result is true, the fault mode is output; otherwise, no fault mode is output. The action selection logic makes a conditional judgment based on the test results and selects a subset of the action set as the output, that is, the subsequent operation to be executed by the current rule knowledge.

[0069] The fault mode needs to include fields such as fault mode name, fault description, equipment node, mechanism analysis, influence scope, and handling suggestions: The fault mode name briefly describes the fault content; the fault description elaborates on the fault content in detail; the equipment node indicates the equipment node to which the fault belongs; the mechanism analysis elaborates on the mechanism of the fault occurrence; the influence scope explains the impact of the fault on equipment functions and mission capabilities, etc.; the handling suggestions give suggestions for subsequent fault handling, including (several) rule knowledge, end, and uncertainty.

[0070] The action set specifies the selection range of subsequent actions for rule knowledge, including (several) rule knowledge, end, and uncertainty. Rule knowledge indicates continuing to execute the corresponding rule knowledge, and there can be multiple. End indicates that a conclusion has been obtained and the fault location process ends normally. Uncertainty indicates that a specific action cannot be selected and the process ends passively.

[0071] To more clearly explain this application, specific examples of this application are provided below.

[0072] As Figure 1 shown is the principle process of rule knowledge diagnosis for ground TT&C equipment.

[0073] Step 1: Obtain the fault phenomenon. The fault phenomenon describes the surface phenomenon of the fault and includes fields such as equipment code, equipment node, and fault description. At the same time, the start and end times for reading data need to be set in this step.

[0074] Step 2: Match the initial rules. The matching conditions are that the equipment code and equipment node are the same and the fault description of the fault phenomenon is similar to the fault description of the rule knowledge fault mode. The matched rule knowledge is the starting point for the inference of the rule knowledge system, and this invention allows multiple paths of the rule knowledge system to execute in parallel.

[0075] Step 3: Initialize the event scheduling center. Initialize the queue of rules knowledge to be inferred as the initially matched rule knowledge, and initialize the event count during execution to 0.

[0076] Step 4: Obtain the next rule knowledge. Take out a rule knowledge from the head of the queue of rules knowledge to be inferred.

[0077] Step 5: Does the rule knowledge exist? If yes, execute the next step; otherwise, jump to Step 15.

[0078] Step 6: Obtain the rule knowledge configuration. When preparing the rule knowledge inference event to be published, read the configuration of the rule knowledge from the database.

[0079] Step 7: Obtain the diagnostic data. Obtain the monitoring points from the rule knowledge configuration and query the diagnostic data from the database in combination with the data start and end times.

[0080] Step 8: Publish the inference event. Put the rule knowledge configuration and diagnostic data into the event and publish the event. At the same time, increment the count of events in execution by one. Jump to step 4.

[0081] Step 9: Asynchronously execute the inference. Perform the rule inference.

[0082] Step 10: After the inference is completed, decrement the count of events in execution by one.

[0083] Step 11: Save the fault mode to the diagnostic conclusion.

[0084] Step 12: Determine the next action? If it is a normal end or an abnormal end, execute the next step; if it is rule knowledge, jump to step 14;

[0085] Step 13: End the process branch.

[0086] Step 14: Put the rule knowledge into the inference queue of the event scheduling center. Jump to step 4.

[0087] Step 21: Is the count zero? If so, jump to step 4; if not, execute the next step.

[0088] Step 16: Output the diagnostic conclusion.

[0089] Step 17: The entire process ends.

[0090] Among them, when performing the rule inference, as Figure 2 shown, it includes the following steps:

[0091] Step 1: Input the time series data and the time slice length. The monitoring point data is collected from the device side and is time series, while the time slice length is a basis for slicing the time series data.

[0092] Step 2: Slice the time series data. Slice the time series data into data segments according to the time slice length, and the data with a length less than one time slice at the end forms a separate data segment.

[0093] Step 3: Traverse the data segments. Until a fault is found.

[0094] Step 4: Conduct traversal tests. All tests must be completed.

[0095] Step 5: Obtain the monitoring point data. According to the monitoring points in the test, retrieve the corresponding data from the data segments.

[0096] Step 6: Feature extraction. For each monitoring point, execute the corresponding feature extraction algorithm to map the time series data into real numbers.

[0097] Step 7: Execute the test algorithm. Input the features of each measurement point into the test algorithm to obtain the test results.

[0098] Step 8: Is the test over? If yes, proceed to the next step; if no, jump to Step 4.

[0099] Step 9: Execute the fault mode judgment logic. Substitute the output of the test into the fault mode judgment logic. If a fault mode occurs, output the fault mode; otherwise, output null.

[0100] Step 10: Execute the action selection logic. Substitute the output of the test into the action selection logic and output the action to continue execution.

[0101] Step 11: Has a fault occurred? If yes, jump to Step 13; if no, proceed to the next step.

[0102] Step 12: Has the traversal of data segments ended? If yes, proceed to the next step; if no, jump to Step 3.

[0103] Step 13: Output the fault mode and action.

[0104] It should be noted that, as Figure 3 shown, the rule knowledge consists of a rule algorithm, fault modes, and an action set. The test is divided into two parts: feature extraction and a test algorithm. The feature extraction algorithm maps time series data into real numbers as the input of the test algorithm. The test algorithm consists of a fault mode judgment logic and an action selection logic, which map the test results to fault modes and actions. The feature extraction algorithm and the test algorithm are saved in the form of drools rule strings in a database table and can be dynamically managed and tested in the algorithm management module.

[0105] As Figure 4 shown, the rule knowledge can form a networked rule knowledge system through the action set. The reasoning process can only proceed along the direction of the arrows, but it does not necessarily traverse the entire knowledge system. It can only pass through some paths or stop halfway along a certain path. To avoid the reasoning process falling into an infinite loop, the present invention prohibits the existence of directed loops in the rule knowledge system.

[0106] As Figure 5 shown, the present application also provides a ground TT&C equipment rule knowledge diagnosis system, including five modules: algorithm management, data interaction, algorithm engine, knowledge reasoning, and diagnosis process.

[0107] The algorithm management module is responsible for adding, deleting, querying, modifying, and testing algorithm scripts. The algorithms include a feature extraction algorithm and a test algorithm, etc., which are stored in a database table in the form of strings, meet the drl file format, allow users to update the algorithms online in real time, and require algorithm testing before updating.

[0108] The data interaction module is used to obtain diagnostic data, test configurations (feature extraction algorithm, test algorithm), and rule knowledge configurations.

[0109] The algorithm engine module (see Figure 6 ) mainly sets the algorithm input, executes the algorithm logic, and obtains the algorithm results, providing support for the feature extraction and test logic of the test. The algorithm engine is implemented based on the drools package. The algorithm input and output are set as global variables of the drl file. The algorithm logic is written in the Java language, and algorithm testing is required before adding or updating to the database. The algorithm's non-running data environment is cached in memory as a KieBase object. The algorithm listener listens in real time for operations such as adding, updating, and deleting algorithms in the database and updates the algorithm environment in memory in real time. The algorithm executor is responsible for executing the algorithm. It first retrieves the algorithm environment from memory and constructs a KieSession algorithm environment with running data. Then, after setting the input data, it executes the algorithm in the KieSession environment, and finally retrieves the algorithm execution results from the KieSession environment.

[0110] The knowledge inference module mainly parses the configuration of knowledge (including the configuration of tests), executes tests, and executes rule algorithms.

[0111] The diagnostic process module mainly provides the diagnostic process, supports interactive diagnostic requirements, and provides functions such as obtaining diagnostic phenomena, matching initial rules, obtaining rule knowledge configurations and diagnostic data, executing rule knowledge inference, scheduling inference events, and confirming diagnostic conclusions.

[0112] In this embodiment, a method and system for rule-based knowledge fault diagnosis of ground measurement and control equipment are provided. By designing rule-based knowledge for ground measurement and control equipment fault diagnosis, dynamic and configurable rule-based knowledge diagnosis is achieved. The diagnosis knowledge update operation is simple, and the diagnosis method can be iteratively upgraded during actual use, solving the technical problems of the current insufficient research on ground measurement and control equipment fault diagnosis methods, inability to adapt to the increasingly tight form of ground measurement and control resources, and inability to meet the urgent PHM needs of users.

[0113] The above content is a further detailed description of the present invention in combination with specific preferred embodiments. It cannot be determined that the specific embodiments of the present invention are limited to this. For those of ordinary skill in the technical field to which the present invention belongs, several simple deductions or substitutions made without departing from the concept of the present invention should be regarded as falling within the patent protection scope determined by the claims submitted by the present invention.

Claims

1. A method for fault diagnosis of regular knowledge of ground measurement and control equipment, characterized in that, the regular knowledge includes tests, rule algorithms, fault modes, and action sets, and the method includes the following steps: S1: Obtain fault phenomena; wherein, the fault phenomena include equipment codes, equipment nodes, and fault descriptions; S2: According to the fault phenomena, match the corresponding initial regular knowledge; S3: Initialize the event scheduling center, initialize the queue of rules knowledge to be inferred as the initial regular knowledge obtained by matching, and initialize the execution event count to 0; S4: Obtain the rules knowledge at the head of the queue of rules knowledge to be inferred; S5: Determine whether the rules knowledge exists. If so, execute step S6; if not, when the execution event count is 0, execute step S4, and when the execution event count is not 0, output the diagnosis conclusion; S6: Obtain the configuration and diagnostic data corresponding to the rules knowledge; S7: Put the configuration and diagnostic data into the event, publish the inference event, increment the execution event count by one, and return to execute step S4; S8: When publishing the inference event, asynchronously execute the rule inference, obtain the inference conclusion, and after the inference is completed, decrement the execution event count by one; wherein, the inference conclusion includes the fault mode and actions; Executing the rule inference specifically includes the following steps: S11: Input the time series data and the length of the time segment; S12: According to the length of the time slice, sequentially divide the time series data into data segments. If the remaining data at the end is less than one segment, it is regarded as one segment; S13: Traverse the data segments; S14: Traverse the tests; S15: According to the monitoring points of the test, extract the data corresponding to the test monitoring points from the data segments; S16: Call the feature extraction algorithm to map the time series data into real numbers; S17: Execute the test algorithm, input the features of each monitoring point into the test algorithm, and obtain the test results; S18: Whether all tests are completed; if so, execute step S19; if not, jump to S14; S19: Execute the fault mode judgment logic, substitute the test results into the fault mode judgment logic, and if the fault mode occurs, output the fault mode; S20: Execute the action selection logic, substitute the test results into the action selection logic, and output the actions to continue execution; S21: Determine whether a fault has occurred. If so, execute step S23; if not, execute step S22; S22: Determine whether the traversal of the data segments is ended. If so, execute step S23; if not, jump to S13; S23: Output the fault mode and actions; S9: If the action is normal end or abnormal end, end the process branch; if the action is regular knowledge, execute step S10; S10: Put the regular knowledge into the queue of rules knowledge to be inferred in the event scheduling center, and execute step S4.

2. The method for fault diagnosis of regular knowledge of ground measurement and control equipment according to claim 1, characterized in that, the tests include a feature extraction algorithm and a test algorithm.

3. The method for fault diagnosis of regular knowledge of ground measurement and control equipment according to claim 2, characterized in that, the feature extraction algorithm is used to map the time series data into real numbers and transmit it as input to the test algorithm.

4. The method for diagnosing faults in regular knowledge of ground measurement and control equipment according to claim 2, characterized in that, the test algorithm is used to map multi-dimensional feature values into Boolean values.

5. The method for diagnosing faults in regular knowledge of ground measurement and control equipment according to claim 2, characterized in that, the feature extraction algorithm and the test algorithm are saved in the form of strings of drools rules to a database table.

6. The method for diagnosing faults in regular knowledge of ground measurement and control equipment according to claim 1, characterized in that, the rule algorithm is a mapping from test results to fault modes and action sets, including fault mode judgment logic and action selection logic.

7. A system for diagnosing faults in regular knowledge of ground measurement and control equipment, characterized in that, it applies to the method for diagnosing faults in regular knowledge of ground measurement and control equipment according to any one of claims 1-6, and the system includes: an algorithm management module for adding, deleting, querying, modifying and testing feature extraction algorithms and test algorithms; a data interaction module for obtaining diagnostic data, test configurations and regular knowledge configurations; an algorithm engine module for setting algorithm inputs, executing algorithm logics and obtaining algorithm results; a knowledge inference module for parsing regular knowledge configurations, executing tests, executing rule algorithms and generating inference conclusions; a diagnostic process module for obtaining diagnostic phenomena, matching initial rules, obtaining regular knowledge configurations and diagnostic data, executing regular knowledge inference, scheduling inference events and confirming diagnostic conclusions.

Images