Mechanical and electrical equipment fault diagnosis method, system and terminal based on rule engine

Abstract

The present invention discloses a method, system and terminal for diagnosing electromechanical equipment faults based on a rule engine, and relates to the technical field of equipment monitoring. The key points of the technical solution are: performing abnormal identification on single-source time series data to obtain an abnormal category set; splitting the abnormal category set into multiple abnormal subsets; calculating the fault estimation probability of each abnormal subset, and arranging the abnormal subsets in descending order according to the fault estimation probability to obtain an abnormal sequence; matching the rule engine package according to the abnormal sequence, and performing fault diagnosis through the rule engine package to obtain a fault diagnosis result and a fault diagnosis probability; when the fault diagnosis probability is lower than a preset threshold, selecting the next sequence number of the abnormal subset for matching diagnosis until the fault diagnosis probability exceeds the preset threshold. The present invention not only takes into account a variety of abnormal situations, but also makes the abnormal diagnosis results more comprehensive and accurate; and through the priority matching of the rule engine package, it preferentially reduces the waste of network resources and improves the efficiency of fault diagnosis.

Description

Technical Field

[0001] The present invention relates to the technical field of equipment monitoring, and more specifically, to a method, system and terminal for diagnosing electromechanical equipment faults based on a rule engine. Background Art

[0002] Equipment failure generally refers to an event or phenomenon in which the equipment loses or reduces its specified function, which is manifested as abnormal equipment production and operation. It means that some parts of the equipment lose their original accuracy or performance, making the equipment unable to operate normally, reducing technical performance, causing equipment interruption or reduced efficiency and affecting production. Large enterprises such as hydropower plants, steel plants, and coal mines are required to conduct health monitoring of the operating status of electromechanical equipment to avoid major economic losses caused by equipment failure.

[0003] At present, the fault diagnosis of electromechanical equipment has evolved from traditional simulation analysis to big data analysis. The main big data fault analysis technology is fault diagnosis and identification based on deep learning neural networks, which uses big data to learn features, thereby characterizing the rich internal information of big data, achieving the fitting of fault data and equipment fault categories, and ultimately improving the accuracy of fault classification or prediction. However, most of the existing electromechanical equipment fault diagnosis is based on the analysis of single-source time series data, and a small part is based on independent analysis after classifying multi-source time series data. In the final analysis, it is all about analyzing a single category of data, and the analysis process does not take into account the correlation between different data. On the one hand, it leads to deviations in the fault diagnosis results, and on the other hand, the diagnosis results are too single and cannot be quickly and accurately analyzed comprehensively.

[0004] Therefore, how to study and design a rule-based engine-based electromechanical equipment fault diagnosis method, system and terminal is a problem that we urgently need to solve. Summary of the invention

[0005] In order to solve the deficiencies in the prior art, the purpose of the present invention is to provide a method, system and terminal for diagnosing electromechanical equipment faults based on a rule engine.

[0006] The above technical objectives of the present invention are achieved through the following technical solutions:

[0007] In a first aspect, a method for diagnosing electromechanical equipment faults based on a rule engine is provided, comprising the following steps:

[0008] Acquire multi-source time series data of the target object, and classify and process the multi-source time series data to obtain multiple independent single-source time series data;

[0009] The single-source time series data is input into the corresponding anomaly recognition model constructed based on the deep neural network for anomaly recognition, and an anomaly category set consisting of multiple single-source categories is obtained;

[0010] Splitting the abnormal category set into multiple abnormal subsets, each abnormal subset includes at least two single-source categories;

[0011] The fault estimation probability of each abnormal subset is calculated according to the correlation of each single source category in the abnormal subset, and the abnormal subsets are arranged in descending order according to the fault estimation probability to obtain an abnormal sequence;

[0012] According to the order of the abnormal sequence, the rule engine package corresponding to the abnormal subset is matched from the database, and the fault diagnosis is performed according to the single-source time series data contained in the abnormal subset through the rule engine package to obtain the fault diagnosis result and fault diagnosis probability;

[0013] When the fault diagnosis probability is lower than the preset threshold, the next abnormal subset with the next sequence number is selected for matching diagnosis until the fault diagnosis probability exceeds the preset threshold, and the final fault diagnosis result is output.

[0014] Furthermore, the calculation process of the fault estimation probability is specifically as follows:

[0015] A single-source category of the anomaly subset is selected as the target category, and the remaining single-source categories are associated categories;

[0016] Determine the probability of the target category occurring when all associated categories occur simultaneously;

[0017] Circularly determine the corresponding occurrence probability of each single-source category in the anomaly subset as the target category to obtain an occurrence probability set;

[0018] The estimated failure probability of the abnormal subset is obtained by processing the occurrence probability set using the mean analysis method.

[0019] Furthermore, the calculation formula of the estimated failure probability is specifically:

[0020] P i n =[P i n (1)+P i n (2)+…+P i n (n)] / n

[0021] Among them, P i n represents the estimated probability of failure corresponding to the abnormal subset i containing n single-source categories; P i n (1) P i n (2) P i n(n) respectively represent the corresponding occurrence probabilities when the first, second, and nth single-source categories in the abnormal subset i are used as the target category.

[0022] Furthermore, the estimated probability of failure of the abnormal subset is calculated as a weighted average value based on the weight coefficient of the corresponding single-source category.

[0023] Furthermore, the weight coefficient of the single-source category is dynamically updated according to the basic factor corresponding to each single-source category in the corresponding abnormal subset.

[0024] Furthermore, the weight coefficient updating process of the single-source category in the abnormal subset is specifically as follows:

[0025] Calculate the sum of the base factors for each single-source category in the anomaly subset;

[0026] The ratio of the basic factor of each single-source category to the sum of the basic factors is analyzed to obtain the dynamic coefficient of the corresponding single-source category;

[0027] The fault estimation probability is calculated by multiplying the dynamic coefficient of the single-source category by the probability of occurrence when the corresponding single-source category is used as the target category.

[0028] Furthermore, the splitting process of the abnormal category set is specifically as follows:

[0029] The number of single-source classes in the anomaly subset of the first split is the same as the number of single-source classes in the anomaly class set;

[0030] The dimensionality reduction is performed step by step by reducing one single-source category at a time until the number of single-source categories in the abnormal subset after the final split is 2.

[0031] Furthermore, if all fault diagnosis probabilities are lower than the preset threshold, the fault estimation probability is multiplied by the fault diagnosis probability to obtain the secondary diagnosis probability of the corresponding abnormal subset, and the fault diagnosis result corresponding to the abnormal subset with the highest secondary diagnosis probability is output as the final fault diagnosis result.

[0032] In the second aspect, a mechanical and electrical equipment fault diagnosis system based on a rule engine is provided, including:

[0033] A data processing module is used to obtain multi-source time series data of the target object, and classify and process the multi-source time series data to obtain multiple independent single-source time series data;

[0034] An anomaly recognition module is used to input the single-source time series data into the corresponding anomaly recognition model constructed based on the deep neural network to perform anomaly recognition and obtain an anomaly category set consisting of multiple single-source categories;

[0035] A category splitting module, used for splitting the abnormal category set into multiple abnormal subsets, each abnormal subset includes at least two single-source categories;

[0036] An estimation and sorting module is used to calculate the fault estimation probability of each abnormal subset according to the correlation of each single source category in the abnormal subset, and arrange the abnormal subsets in descending order according to the fault estimation probability to obtain an abnormal sequence;

[0037] A matching diagnosis module is used to match the rule engine package of the corresponding abnormal subset from the database according to the order of the abnormal sequence, and perform fault diagnosis based on the single-source time series data contained in the abnormal subset through the rule engine package to obtain the fault diagnosis result and fault diagnosis probability;

[0038] The loop control module is used to select the next abnormal subset with a sequence number for matching diagnosis when the fault diagnosis probability is lower than the preset threshold, until the fault diagnosis probability exceeds the preset threshold, and output the final fault diagnosis result.

[0039] In a third aspect, a computer terminal is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, a rule engine-based electromechanical equipment fault diagnosis method as described in any one of the first aspects is implemented.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] 1. The present invention analyzes the abnormal correlation between different categories, conducts a comprehensive analysis on multiple abnormal data appearing at the same time, prioritizes possible fault conditions with higher probability according to probability sorting, and analyzes the corresponding fault conditions by matching the corresponding rule engine packages; on the one hand, multiple abnormal conditions are considered, making the abnormal diagnosis results more comprehensive and accurate; on the other hand, through the priority matching of the rule engine package, the waste of network resources is preferentially reduced, and the efficiency of fault diagnosis is improved;

[0042] 2. The present invention dynamically updates the weight coefficient of each single-source category according to the abnormal category in the abnormal subset, which not only takes into account the severity of the faults corresponding to different unit categories, but also can adapt to the abnormal subsets composed of different single-source categories, thereby improving the calculation accuracy of the fault estimation probability of different abnormal subsets. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The drawings described herein are used to provide a further understanding of the embodiments of the present invention, constitute a part of this application, and do not constitute a limitation of the embodiments of the present invention. In the drawings:

[0044] Figure 1 is a flow chart in an embodiment of the present invention;

[0045] Figure 2 is a system block diagram of an embodiment of the present invention. DETAILED DESCRIPTION

[0046] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with embodiments and drawings. The exemplary embodiments of the present invention and their description are only used to explain the present invention and are not intended to limit the present invention.

[0047] Example 1: A method for diagnosing electromechanical equipment faults based on a rule engine, such as Figure 1 As shown, the following steps are included:

[0048] S1: Obtain multi-source time series data of the target object, and classify and process the multi-source time series data to obtain multiple independent single-source time series data;

[0049] S2: Input the single-source time series data into the corresponding anomaly recognition model built based on the deep neural network for anomaly recognition, and obtain an anomaly category set consisting of multiple single-source categories;

[0050] S3: Split the abnormal category set into multiple abnormal subsets, each of which includes at least two single-source categories;

[0051] S4: Calculate the fault estimation probability of each abnormal subset according to the correlation of each single source category in the abnormal subset, and arrange the abnormal subsets in descending order according to the fault estimation probability to obtain an abnormal sequence;

[0052] S5: matching the rule engine package corresponding to the abnormal subset from the database according to the sequence of the abnormal sequence, and performing fault diagnosis based on the single-source time series data contained in the abnormal subset through the rule engine package to obtain the fault diagnosis result and fault diagnosis probability;

[0053] S6: When the fault diagnosis probability is lower than the preset threshold, the next sequence numbered abnormal subset is selected for matching diagnosis until the fault diagnosis probability exceeds the preset threshold, and the final fault diagnosis result is output.

[0054] In step S4, the calculation process of the estimated fault probability is specifically as follows:

[0055] S41: Select a single-source category of the abnormal subset as the target category, and the remaining single-source categories as associated categories;

[0056] S41: Determine the probability of the target category occurring when all associated categories occur simultaneously;

[0057] S41: cyclically determine the occurrence probability corresponding to each single-source category in the abnormal subset as the target category, and obtain an occurrence probability set;

[0058] S41: After processing the occurrence probability set by the mean analysis method, the estimated failure probability of the abnormal subset is obtained.

[0059] The calculation formula for the estimated probability of failure is as follows:

[0060] P i n =[P i n (1)+P i n (2)+…+P i n (n)] / n

[0061] Among them, P i n represents the estimated probability of failure corresponding to the abnormal subset i containing n single-source categories; P i n (1) P i n (2) P i n (n) respectively represent the corresponding occurrence probabilities when the first, second, and nth single-source categories in the abnormal subset i are used as the target category.

[0062] In addition, in order to consider the severity of the faults corresponding to different unit categories, the estimated fault probability of the abnormal subset can also be calculated as a weighted average value based on the weight coefficient of the corresponding single-source category.

[0063] In addition, in order to adapt to the abnormal subsets composed of different single-source categories and improve the calculation accuracy of the fault estimation probability of different abnormal subsets, the weight coefficient of the single-source category can be dynamically updated according to the basic factor corresponding to each single-source category in the corresponding abnormal subset.

[0064] The weight coefficient updating process of the single-source category in the abnormal subset is as follows: calculate the sum of the basic factors of each single-source category in the abnormal subset; analyze and obtain the ratio of the basic factor of each single-source category to the sum of the basic factors to obtain the dynamic coefficient of the corresponding single-source category; multiply the dynamic coefficient of the single-source category by the probability of occurrence when the corresponding single-source category is used as the target category to realize the fault estimation probability calculation.

[0065] The specific process of splitting the abnormal category set is as follows: the number of single-source categories in the abnormal subset of the first split is the same as the number of single-source categories in the abnormal category set; the dimensionality reduction split is performed step by step by reducing one single-source category each time until the number of single-source categories in the abnormal subset of the final split is 2.

[0066] It should be noted that in order to ensure that each diagnostic process has a final output result and avoid the situation where there is no output result due to the preset threshold being set too high, if all fault diagnosis probabilities are lower than the preset threshold, the fault estimation probability is multiplied by the fault diagnosis probability to obtain the secondary diagnosis probability of the corresponding abnormal subset, and the fault diagnosis result corresponding to the abnormal subset with the highest secondary diagnosis probability is output as the final fault diagnosis result.

[0067] Example 2: Mechanical and electrical equipment fault diagnosis system based on rule engine, such as Figure 2 As shown, it includes a data processing module, an anomaly recognition module, a category splitting module, an estimation and sorting module, a matching diagnosis module and a loop control module.

[0068] Among them, the data processing module is used to obtain the multi-source time series data of the target object, and obtain multiple independent single-source time series data after classifying the multi-source time series data. The abnormal identification module is used to input the single-source time series data into the corresponding abnormal identification model constructed according to the deep neural network for abnormal identification, and obtain an abnormal category set composed of multiple single-source categories. The category splitting module is used to split the abnormal category set into multiple abnormal subsets, each of which includes at least two single-source categories. The estimation sorting module is used to calculate the fault estimation probability of each abnormal subset according to the correlation of each single-source category in the abnormal subset, and arrange the abnormal subsets in descending order according to the fault estimation probability to obtain an abnormal sequence. The matching diagnosis module is used to match the rule engine package of the corresponding abnormal subset from the database according to the order of the abnormal sequence, and perform fault diagnosis based on the single-source time series data contained in the abnormal subset through the rule engine package to obtain the fault diagnosis result and the fault diagnosis probability. The loop control module is used to select the next abnormal subset with a sequence number for matching diagnosis when the fault diagnosis probability is lower than the preset threshold, until the fault diagnosis probability exceeds the preset threshold, and output the final fault diagnosis result.

[0069] Working principle: The present invention analyzes the abnormal correlation between different categories, conducts a comprehensive analysis of various abnormal data appearing at the same time, prioritizes possible fault conditions with higher probabilities according to probability sorting, and analyzes the corresponding fault conditions by matching the corresponding rule engine packages; on the one hand, it takes into account multiple abnormal conditions, making the abnormal diagnosis results more comprehensive and accurate; on the other hand, through the priority matching of the rule engine package, it prioritizes the reduction of network resource waste and improves the efficiency of fault diagnosis.

[0070] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0071] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0072] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0073] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0074] The above specific implementation methods further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above are only specific implementation methods of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for diagnosing electromechanical equipment faults based on a rule engine, characterized in that: The following steps are involved: Acquire multi-source time series data of the target object, and classify and process the multi-source time series data to obtain multiple independent single-source time series data; The single-source time series data is input into the corresponding anomaly recognition model constructed based on the deep neural network for anomaly recognition, and an anomaly category set consisting of multiple single-source categories is obtained; Splitting the abnormal category set into multiple abnormal subsets, each abnormal subset includes at least two single-source categories; The fault estimation probability of each abnormal subset is calculated according to the correlation of each single source category in the abnormal subset, and the abnormal subsets are arranged in descending order according to the fault estimation probability to obtain an abnormal sequence; According to the order of the abnormal sequence, the rule engine package corresponding to the abnormal subset is matched from the database, and the fault diagnosis is performed according to the single-source time series data contained in the abnormal subset through the rule engine package to obtain the fault diagnosis result and fault diagnosis probability; When the fault diagnosis probability is lower than the preset threshold, the next sequence numbered abnormal subset is selected for matching diagnosis until the fault diagnosis probability exceeds the preset threshold, and the final fault diagnosis result is output.

2. The electromechanical equipment fault diagnosis method based on rule engine according to claim 1 is characterized in that: The calculation process of the fault estimation probability is specifically as follows: A single-source category of the anomaly subset is selected as the target category, and the remaining single-source categories are associated categories; Determine the probability of the target category occurring when all associated categories occur simultaneously; Circularly determine the corresponding occurrence probability of each single-source category in the anomaly subset as the target category to obtain an occurrence probability set; The estimated failure probability of the abnormal subset is obtained by processing the occurrence probability set using the mean analysis method.

3. The electromechanical equipment fault diagnosis method based on rule engine according to claim 2 is characterized in that: The calculation formula of the failure estimation probability is specifically: P i n =[P i n (1)+P i n (2)+…+P i n (n)] / n Among them, P i n represents the estimated probability of failure corresponding to the abnormal subset i containing n single-source categories; P i n (1) P i n (2) P i n (n) respectively represent the corresponding occurrence probabilities when the first, second, and nth single-source categories in the abnormal subset i are used as the target category.

4. The electromechanical equipment fault diagnosis method based on rule engine according to claim 2 is characterized in that: The estimated probability of failure of the abnormal subset is calculated by weighted average value based on the weight coefficient of the corresponding single source category.

5. The electromechanical equipment fault diagnosis method based on rule engine according to claim 4 is characterized in that: The weight coefficient of the single-source category is dynamically updated according to the basic factor corresponding to each single-source category in the corresponding abnormal subset.

6. The electromechanical equipment fault diagnosis method based on rule engine according to claim 5 is characterized in that: The weight coefficient updating process of the single-source category in the abnormal subset is specifically as follows: Calculate the sum of the base factors for each single-source category in the anomaly subset; The ratio of the basic factor of each single-source category to the sum of the basic factors is analyzed to obtain the dynamic coefficient of the corresponding single-source category; The fault estimation probability is calculated by multiplying the dynamic coefficient of the single-source category by the probability of occurrence when the corresponding single-source category is used as the target category.

7. The electromechanical equipment fault diagnosis method based on rule engine according to claim 1 is characterized in that: The splitting process of the abnormal category set is specifically as follows: The number of single-source classes in the anomaly subset of the first split is the same as the number of single-source classes in the anomaly class set; The dimensionality reduction is performed step by step by reducing one single-source category at a time until the number of single-source categories in the abnormal subset after the final split is 2.

8. The electromechanical equipment fault diagnosis method based on rule engine according to any one of claims 1 to 7, characterized in that: If all fault diagnosis probabilities are lower than the preset threshold, the fault estimation probability is multiplied by the fault diagnosis probability to obtain the secondary diagnosis probability of the corresponding abnormal subset, and the fault diagnosis result corresponding to the abnormal subset with the highest secondary diagnosis probability is output as the final fault diagnosis result.

9. The electromechanical equipment fault diagnosis system based on rule engine is characterized by: include: A data processing module is used to obtain multi-source time series data of the target object, and classify and process the multi-source time series data to obtain multiple independent single-source time series data; An anomaly recognition module is used to input the single-source time series data into the corresponding anomaly recognition model constructed based on the deep neural network to perform anomaly recognition and obtain an anomaly category set consisting of multiple single-source categories; A category splitting module, used for splitting the abnormal category set into multiple abnormal subsets, each abnormal subset includes at least two single-source categories; An estimation and sorting module is used to calculate the fault estimation probability of each abnormal subset according to the correlation of each single source category in the abnormal subset, and arrange the abnormal subsets in descending order according to the fault estimation probability to obtain an abnormal sequence; A matching diagnosis module is used to match the rule engine package of the corresponding abnormal subset from the database according to the order of the abnormal sequence, and perform fault diagnosis based on the single-source time series data contained in the abnormal subset through the rule engine package to obtain the fault diagnosis result and fault diagnosis probability; The loop control module is used to select the next abnormal subset with a sequence number for matching diagnosis when the fault diagnosis probability is lower than the preset threshold, until the fault diagnosis probability exceeds the preset threshold, and output the final fault diagnosis result.

10. A computer terminal comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the electromechanical equipment fault diagnosis method based on the rule engine as described in any one of claims 1 to 8 is implemented.

Images