Edge cloud collaborative equipment fault diagnosis method and device

Abstract

The invention relates to the technical field of equipment health management, and provides an edge cloud collaborative equipment fault diagnosis method and device, which can realize quick response of intelligent operation and maintenance business of high-end equipment. According to the invention, the fault diagnosis algorithm is cooperatively operated on the edge device and the cloud platform, the data acquisition, local data processing and fault diagnosis capabilities of edge computing are utilized, and the data analysis, data sharing and comprehensive fault diagnosis capabilities of the cloud platform are fully utilized, so that the quick response of the intelligent operation and maintenance service of the high-end device is realized.

Description

Technical Field

[0001] The present invention belongs to the technical field of equipment health management, and specifically relates to an edge-cloud collaborative equipment fault diagnosis method and device. Background Art

[0002] With the rapid development of high-end equipment such as aerospace, aviation, high-speed rail, energy, and large facilities in my country, the country has put forward higher requirements for the safety, economy, and intelligence of high-end equipment. In order to meet the above needs of high-end equipment operation and security, health management technology is gradually gaining attention. Health management technology (hereinafter referred to as PHM technology) is an advanced management technology based on big data, which realizes the intelligence and autonomy of equipment status through real-time monitoring, diagnosis, and prediction of equipment through intelligent processing algorithms. It is the core technology to improve the safety and economy of equipment.

[0003] The key content of PHM is the data collection and analysis of high-end equipment on site. The realization of efficient collection and real-time processing of equipment data is the primary condition for completing equipment working status monitoring and data analysis. However, if most of the fault diagnosis and analysis work is submitted to the cloud for processing, the massive amount of data collected on site will not be able to take into account the data collection efficiency and the real-time requirements of fault diagnosis. It can be seen that the current equipment fault detection technology cannot meet the needs of real-time fault diagnosis and edge-cloud collaborative data processing. Summary of the invention

[0004] In view of the above problems, the present invention proposes to provide an edge-cloud collaborative equipment fault diagnosis method and device, which can achieve rapid response of high-end equipment intelligent operation and maintenance services.

[0005] To achieve the above purpose, the technical solution of the present invention is as follows:

[0006] An edge-cloud collaborative equipment fault diagnosis method, characterized by comprising the following steps:

[0007] Compile fault diagnosis knowledge for the target object in the cloud, wherein the fault diagnosis knowledge at least includes rules and algorithms;

[0008] Package the compiled fault diagnosis knowledge and send it to the edge device through the communication protocol;

[0009] The edge device receives and parses the fault diagnosis knowledge package and verifies it. If the verification passes, it proceeds to the subsequent stage. Otherwise, it continues to communicate with the cloud until the correct fault diagnosis knowledge package is obtained.

[0010] The edge device updates and loads fault diagnosis knowledge and restarts the software in idle state;

[0011] Start the data collection process. The edge device parses and stores the data according to the data collection protocol, and sends the characteristic data to the cloud according to the data characteristics.

[0012] The cloud receives and parses data from edge devices and stores it in the cloud database;

[0013] The edge device starts the fault diagnosis process, makes a fault diagnosis based on the fault knowledge, and sends the fault information to the cloud if a fault is found;

[0014] According to the fault situation, the fault location process is started. If the fault can be located, troubleshooting measures are prompted. Otherwise, the original data is sent to the cloud for comprehensive diagnosis.

[0015] If the cloud can locate the fault, it will prompt troubleshooting measures, otherwise it will inform the user that it cannot be located.

[0016] The communication protocol defines a synchronization method between the two parties and a check bit of fault diagnosis knowledge.

[0017] The data acquisition protocol describes the communication protocol, message start identifier, message length, message type, device acquisition ID, device acquisition channel, message content and verification content used in the communication process.

[0018] Among them, for rapidly changing data, key time-frequency domain features including time-domain waveform pulse factor and frequency-domain amplitude skewness are extracted and sent as feature data.

[0019] Among them, the fault information sent by the edge device to the cloud after discovering a fault includes an existing fault code or an unknown fault, where the fault code number is represented by 2 bytes.

[0020] Among them, when the edge device is unable to perform fault diagnosis, the cloud sends command information to the edge device, and the edge device responds to the command and packages the original data and sends it to the cloud.

[0021] Among them, when the cloud performs comprehensive diagnosis on the original data, the methods used include but are not limited to machine learning methods based on historical data and big data, mechanism-based fault diagnosis methods, and rule-based fault diagnosis methods.

[0022] The present invention also provides an edge-cloud collaborative equipment fault diagnosis device for implementing the method of the present invention, characterized in that it comprises: an edge device device and a cloud device, wherein the edge device device comprises: a data access unit for providing an interface with an external hardware system, a data parsing unit for performing data parsing according to an acquisition communication protocol of a sensor, a feature extraction unit for extracting feature values ​​from raw data, a fault diagnosis unit for interpreting data based on an inference engine to obtain a diagnosis result, and a data forwarding unit for the edge device to send data, fault diagnosis results, and receive data instructions from the cloud;

[0023] The cloud device includes: a data access unit for collecting data information from multiple edge devices and sending different types of data to corresponding modules for processing, an information alarm unit for prompting users about fault information, a comprehensive diagnosis unit for performing comprehensive analysis in combination with historical data and multiple fault diagnosis methods, an information storage unit for storing data received from the data access unit according to predetermined rules, and an auxiliary maintenance unit for providing auxiliary maintenance suggestions and helping users to query and locate relevant maintenance measures.

[0024] Among them, the information alarm unit of the cloud device prompts the user with fault information in a visual manner, including the fault name, number and processing urgency.

[0025] The beneficial effects of the present invention are as follows:

[0026] 1. The fault diagnosis algorithm adopted in the present invention runs in coordination between the edge device and the cloud platform, which not only utilizes the data collection, local data processing and fault diagnosis capabilities of edge computing, but also makes full use of the data analysis, data sharing and comprehensive fault diagnosis capabilities of the cloud platform, thereby achieving rapid response to the intelligent operation and maintenance services of high-end equipment.

[0027] 2. In the present invention, the fault diagnosis knowledge compiled in the cloud adopts heuristic rules and multiple algorithms to ensure the scientificity and accuracy of the diagnosis basis. The edge device analyzes and verifies the fault diagnosis knowledge package, further ensuring the accuracy of the knowledge, thereby improving the accuracy of fault diagnosis.

[0028] 3. In the present invention, a series of processes including knowledge compilation, packaging, distribution, synchronization, analysis and verification in the pre-diagnosis preparation stage, and data collection, analysis, storage, sending, receiving, fault diagnosis, and positioning in the diagnosis implementation stage form a complete and efficient diagnostic system and optimize the entire diagnostic process.

[0029] 4. In the present invention, during the fault diagnosis process, accurate judgment and location can be made according to the fault situation. For faults that can be located, troubleshooting measures can be prompted in a timely manner; for faults that cannot be located, comprehensive diagnosis is performed through the collaboration of the cloud and edge devices, thereby enhancing the ability to handle various types of faults.

[0030] 5. In the present invention, efficient communication and collaboration between cloud and edge devices are achieved through edge-cloud collaboration. The improvement of communication protocols ensures the accuracy and reliability of data transmission. Each module has a clear division of labor and works in collaboration, which improves the overall collaborative efficiency. It effectively improves the fault diagnosis capability and response speed of high-end equipment, timely discovers and solves faults, reduces equipment losses caused by faults, and extends the service life of equipment.

[0031] 6. The device of the present invention uses edge device devices and cloud devices, combined with fault diagnosis algorithms, to work together on edge devices and cloud platforms. It not only utilizes the data collection, local data processing, and fault diagnosis capabilities of edge computing, but also fully utilizes the data analysis, data sharing, and comprehensive fault diagnosis capabilities of the cloud platform, thereby achieving rapid response to intelligent operation and maintenance services for high-end equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is a flow chart of an edge-cloud collaborative equipment fault diagnosis method according to an embodiment of the present invention.

[0033] Figure 2 It is a flow chart of the communication protocol execution required for the fault diagnosis of the embodiment of the present invention.

[0034] Figure 3 It is a structural schematic diagram of an edge-cloud collaborative equipment fault diagnosis device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0035] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0036] The present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0037] The present invention provides an edge-cloud collaborative equipment fault diagnosis method. Figure 1: is a flow chart of an edge-cloud collaborative equipment fault diagnosis method according to an embodiment of the present invention. An edge-cloud collaborative equipment fault diagnosis method according to an embodiment of the present invention comprises the following steps:

[0038] 1) Pre-diagnosis preparation stage

[0039] (1) In the cloud, fault diagnosis knowledge (rules and algorithms) is compiled for the target object; the fault diagnosis knowledge can be described based on rules and algorithms, where the rules are described using heuristic rules; the algorithms can be some feature extraction, data filtering and other methods;

[0040] (2) After the compilation is completed, the fault diagnosis knowledge is packaged and sent to the edge device through the communication protocol;

[0041] (3) After receiving the fault diagnosis knowledge package, the edge device synchronizes the two parties according to the communication protocol and parses the fault diagnosis knowledge package. To ensure that there is no problem with the fault diagnosis knowledge, the fault diagnosis knowledge package is verified. If the verification passes, it proceeds to the subsequent stage. If it fails, it continues to communicate with the cloud to obtain the fault diagnosis knowledge package.

[0042] Preferably: the communication protocol is a system predefined protocol content, and the protocol also defines the content of the synchronization method between the two parties;

[0043] Preferably: in order to ensure that there is no error in the transmission process of the fault diagnosis knowledge transmitted from the cloud to the edge device, the fault diagnosis knowledge related check bits are defined in the protocol, and the edge terminal will check the received fault diagnosis knowledge package. If the check passes, it will enter the subsequent link. If it fails, it will continue to communicate with the cloud to obtain the fault diagnosis knowledge package;

[0044] (4) The edge device waits for it to enter the idle state (complete the processing of communication data once). If the edge device is in a busy state, it will wait in unison. Otherwise, it will update and load the fault knowledge and restart the software. That is, the edge device will update the data only when it is in the idle state. Specifically, the edge device determines whether it is in the idle state based on the data collection and processing flow. If it completes a data collection and processing flow, it will stop collecting. During this process, the system will wait. After completion, it will update and load the fault knowledge and restart the edge device software.

[0045] 2) Diagnosis implementation stage

[0046] (5) Start the data collection process;

[0047] (6) The edge device parses and stores the data according to the data acquisition protocol, and sends the data according to the characteristics of the data (such as fast-changing data and slow-changing data, characteristic data); the data acquisition protocol describes the communication protocol (TCP, UDP, etc.) used in the communication process, the message start identifier, the message length, the message type, the device acquisition ID, the device acquisition channel, the message content, the verification content, etc. For different signal contents, the signal characteristics are processed according to the signal characteristics, including fast-changing data, slow-changing data, etc. At the same time, for fast-changing data, its characteristic data is extracted, including the time domain waveform pulse factor, the frequency domain amplitude skewness and other time-frequency domain key features; after the characteristic data is extracted, the data can be sent according to the communication protocol;

[0048] (7) Receive and analyze data according to the data communication protocol and store it in the cloud database;

[0049] (8) The edge device starts the fault diagnosis process and makes a fault diagnosis based on the fault knowledge. If a fault occurs, the fault information is sent to the cloud according to the communication protocol; otherwise, the fault information is continued to be determined;

[0050] Among them, for fault diagnosis, the edge terminal makes a fault judgment based on the fault knowledge base. If a fault is found, the fault code information or unknown fault information is sent to the cloud according to the communication protocol. The cloud can issue a warning and assist in the positioning process.

[0051] (9) Start the fault location process according to the fault situation. If the fault can be located, go to step (12); otherwise, go to step (10);

[0052] (10) The cloud system sends command information to the edge device. After receiving the command, the edge device recognizes the command and packages the original data and sends it to the cloud.

[0053] (11) The cloud software combines fault diagnosis methods to conduct comprehensive diagnosis of the original data;

[0054] (12) If the fault can be located, the fault location can be performed and troubleshooting measures can be prompted. If the fault cannot be located, the user will be prompted that the fault cannot be located.

[0055] Preferably: during the fault location process, the cloud can perform fault location and repair and troubleshooting based on the feedback fault code, and the system can perform fault location and repair and troubleshooting based on the fault code information. If the feedback is an unknown fault, the cloud can send instruction information to the edge device. After receiving the instruction, the edge device recognizes the instruction, packages the original data and sends it to the cloud. With the help of the original data, comprehensive cloud diagnosis is performed, including machine learning methods based on historical data and big data, fault diagnosis methods based on mechanisms, fault diagnosis methods based on rules, and other methods. If the fault can be located, the fault can be located and troubleshooting measures can be prompted. If the fault cannot be located, the user will be prompted that the fault cannot be located.

[0056] The above technical solution of the embodiment of the present invention is described in detail below in conjunction with the accompanying drawings.

[0057] exist Figure 2 The communication protocol execution flow chart required for fault diagnosis in the method of the present invention is given in the figure. The communication protocol mainly includes a message start mark, a message end mark, a message level, a message type, a timestamp, a message content, a check, etc. The main message content is as follows:

[0058] 1) After the system is started, the edge device will register with the cloud. The registration protocol includes the device number, board number, board type, and board number of the registered edge device;

[0059] 2) After the fault diagnosis knowledge is compiled, the cloud can send the fault diagnosis knowledge to the edge device. The fault diagnosis knowledge protocol mainly includes the knowledge rule content and the verification code of the knowledge rule (the verification code can be generated by the md5 method);

[0060] 3) After data collection, the edge device can send data information to the cloud. The data information includes characteristic data of fast-changing data and slow-changing data. The characteristic data includes time domain waveform mean, time domain waveform root mean square, center of gravity frequency, mean square frequency, frequency domain amplitude skewness index and other data contents. Each characteristic data can be represented in byte form (for example, 4 bytes); for slow-changing data, the original data can be uploaded;

[0061] 4) After the edge device diagnoses, if a fault is found, it sends fault information to the cloud, including existing fault codes or unknown faults. The main communication protocol content includes the fault code number. The fault code number can be identified by 2 bytes, from 0000 to FFFF, where FFFF indicates an unknown fault.

[0062] 5) When the edge device is unable to perform fault diagnosis, the cloud sends raw data extraction instructions to the edge device. The main protocol is data extraction instructions, data extraction start time, data extraction end time, etc.;

[0063] 6) The edge device sends the original data to the cloud device. For the original data that changes rapidly, the edge device can package the original data into a compressed file and transmit it to the cloud via the FTP protocol.

[0064] The fault diagnosis knowledge representation is as follows:

[0065] Fault diagnosis knowledge defines the grammatical system of knowledge representation based on the condition-expected value interpretation method, and stipulates the keywords, operators, statement formats, etc. of knowledge representation:

[0066] 1) Keywords supported by knowledge representation

[0067] Parameter attribute keywords. Including: N (original code), U (voltage), Y (physical quantity), EV (expected value), MAX (maximum value operation), MIN (minimum value operation), DELTA (difference operation), ERR (error limit), this (knowledge owner), etc.

[0068] Conclusion keywords. Including: true (conclusion is correct), false (conclusion is wrong), etc.

[0069] Statement keywords. Including conditional statement keywords (if then else), assignment statement keywords (=), etc.

[0070] Operator keywords include mathematical operators (+-* / ), relational operators (==><>=<=!=), logical operators (&||!&|), etc.

[0071] Function keywords, including trigonometric functions, change law functions, and other functions.

[0072] 2) Supported data types

[0073] The supported data types vary depending on the data in the knowledge representation. For test data engineering values, floating point types are supported; for state quantity information, instruction sending description information, state information, etc. in the test data, integer types are supported.

[0074] 3) Supported statements and operations

[0075] Statements include assignment statements, conditional statements, and comment statements. Operations include relational operations, logical operations, and mathematical operations.

[0076] exist Figure 3An edge-cloud collaborative equipment fault diagnosis device of the present embodiment is given, including: an edge device device (data access unit 60, a data analysis unit 62, a feature extraction unit 64, a fault diagnosis unit 66 and a data forwarding unit 68) and a cloud device (data access unit 70, an information alarm unit 74, a comprehensive diagnosis unit 72, an information storage unit 76 and an auxiliary maintenance unit 78). The following is a detailed description of each module of the embodiment of the present invention.

[0077] Edge Devices:

[0078] Data access unit 60: implements access to different external sensor interfaces, and allocates data memory for access, and performs data storage and exchange. Specifically used for: providing an interface with an external hardware system (such as a sensor interface), etc., and allowing data communication transmission through a communication protocol (MODBUS, TCP, etc.) according to configuration information;

[0079] Data analysis unit 62: performs data analysis according to the sensor's acquisition communication protocol, mainly including data de-framing, data conversion (conversion from original value to engineering value), etc.;

[0080] Feature extraction unit 64: transforms data feature values ​​according to the original data, including time domain data, frequency domain data, and other contents, such as time domain waveform mean, time domain waveform root mean square, centroid frequency, mean square frequency, frequency domain amplitude skewness index, and other contents;

[0081] Fault diagnosis unit 66: The fault diagnosis unit loads knowledge, interprets data, etc. according to preset fault diagnosis knowledge. The diagnosis works through a data-driven reasoning diagnosis system and is based on an inference engine. The inference engine adopts a specific reasoning strategy, method and inference algorithm, and realizes the reasoning process through the rules in the knowledge base loaded in the blackboard area to obtain the result of the diagnosis interpretation.

[0082] Data forwarding unit 68: The data forwarding unit sends data, fault diagnosis results, and receiving data instructions to the cloud for the edge device;

[0083] Cloud Devices:

[0084] Data access unit 70: collects data information from multiple edge devices, and sends different data contents to different modules; sends data information to the information storage unit; sends fault information to the information alarm unit and the comprehensive diagnosis unit; at the same time, the data access unit can also send fault diagnosis knowledge, control instructions and other contents to the edge devices.

[0085] Comprehensive diagnosis unit 72: If the fault information sent on-site cannot be directly used for fault location, the comprehensive diagnosis unit receives all data information, and checks the data change trend based on the fault diagnosis rule method, the mechanism-based fault diagnosis method, and the big data-based fault diagnosis method, especially in combination with historical data information, so as to locate the fault according to the comprehensive diagnosis plan;

[0086] Information alarm unit 74: Prompts the user with fault information in a visual manner, and the user can view the name, number, and urgency of the fault through the fault information;

[0087] Information storage unit 76: after receiving data from the data access unit, stores the data according to a predetermined data rule;

[0088] Auxiliary maintenance unit 78: provides auxiliary maintenance suggestions. According to the fault code, relevant maintenance suggestion content can be queried and located.

[0089] In summary, with the help of the technical solution of the embodiment of the present invention, through the edge device device and the cloud device, the fault diagnosis algorithm is run in coordination between the edge device and the cloud platform, which not only utilizes the data collection, local data processing and fault diagnosis capabilities of edge computing, but also makes full use of the data analysis, data sharing and comprehensive fault diagnosis capabilities of the cloud platform, thereby achieving rapid response to the intelligent operation and maintenance business of high-end equipment.

[0090] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. An edge-cloud collaborative equipment fault diagnosis method, characterized in that: The following steps are involved: Compile fault diagnosis knowledge for the target object in the cloud, wherein the fault diagnosis knowledge at least includes rules and algorithms; Package the compiled fault diagnosis knowledge and send it to the edge device through the communication protocol; The edge device receives and parses the fault diagnosis knowledge package and verifies it. If the verification passes, it proceeds to the subsequent stage. Otherwise, it continues to communicate with the cloud until the correct fault diagnosis knowledge package is obtained. The edge device updates and loads fault diagnosis knowledge and restarts the software in idle state; Start the data collection process. The edge device parses and stores the data according to the data collection protocol, and sends the characteristic data to the cloud according to the data characteristics. The cloud receives and parses data from edge devices and stores it in the cloud database; The edge device starts the fault diagnosis process, makes a fault diagnosis based on the fault knowledge, and sends the fault information to the cloud if a fault is found; According to the fault situation, the fault location process is started. If the fault can be located, troubleshooting measures are prompted. Otherwise, the original data is sent to the cloud for comprehensive diagnosis. If the cloud can locate the fault, it will prompt troubleshooting measures, otherwise it will inform the user that it cannot be located.

2. The method according to claim 1, characterized in that The communication protocol defines a method for synchronization between the two parties and a check bit for fault diagnosis knowledge.

3. The method according to claim 1 or 2, characterized in that The data acquisition protocol describes the communication protocol, message start identifier, message length, message type, device acquisition ID, device acquisition channel, message content and verification content used in the communication process.

4. The method according to claim 3, characterized in that For fast-changing data, key time-frequency domain features including time-domain waveform pulse factor and frequency-domain amplitude skewness are extracted and sent as feature data.

5. The method according to claim 1, characterized in that After discovering a fault, the edge device sends the fault information to the cloud, including an existing fault code or an unknown fault, where the fault code number is represented by 2 bytes.

6. The method according to claim 4 or 5, characterized in that When the edge device is unable to perform fault diagnosis, the cloud sends command information to the edge device, and the edge device responds to the command and packages the original data and sends it to the cloud.

7. The method according to claim 6, characterized in that When the cloud performs comprehensive diagnosis on raw data, the methods used include but are not limited to machine learning methods based on historical data and big data, mechanism-based fault diagnosis methods, and rule-based fault diagnosis methods.

8. An edge-cloud collaborative equipment fault diagnosis device for implementing the method according to any one of claims 1 to 7, characterized in that: include: An edge device and a cloud device, wherein the edge device comprises: a data access unit for providing an interface with an external hardware system, a data parsing unit for performing data parsing according to a sensor acquisition communication protocol, a feature extraction unit for extracting feature values ​​from raw data, a fault diagnosis unit for interpreting data based on an inference engine to obtain a diagnosis result, and a data forwarding unit for the edge device to send data, fault diagnosis results, and receive data instructions from the cloud; The cloud device includes: a data access unit for collecting data information from multiple edge devices and sending different types of data to corresponding modules for processing, an information alarm unit for prompting users about fault information, a comprehensive diagnosis unit for performing comprehensive analysis in combination with historical data and multiple fault diagnosis methods, an information storage unit for storing data received from the data access unit according to predetermined rules, and an auxiliary maintenance unit for providing auxiliary maintenance suggestions and helping users to query and locate relevant maintenance measures.

9. The device according to claim 8, characterized in that The information alarm unit of the cloud device prompts the user with fault information in a visual manner, including the fault name, number and urgency of handling.

Images