An MAS-based Adaptive Device Health Diagnosis System and Diagnosis Method

Abstract

The present invention discloses an MAS-based adaptive device health diagnosis system and a diagnosis method, which relate to the technical field of device predictive maintenance. The diagnosis system is a single-layer multi-agent system composed of N diagnostic intelligent terminals, and any two of the N diagnostic intelligent terminals are connected by wireless communication or wired connection. The diagnosis system provided by the present invention has high self-adaptability and flexibility, can independently complete multi-layer target tasks such as multi-dimensional information perception, data processing, fusion judgment, and AI self-adaptation. The intelligent agent products can automatically handshake and cooperate to complete complex tasks. Users do not need to perform front-end development, and the development complexity and workload of solution providers are greatly reduced. The addition or withdrawal of business nodes, and the increase, decrease, and adjustment of local logic functions do not require rewriting the system software.

Description

Technical Field

[0001] The present invention relates to the technical field of predictive maintenance of equipment, and particularly to an adaptive equipment health diagnosis system and diagnosis method based on MAS. Background Art

[0002] With the advent of the Industry 4.0 era, equipment operation and maintenance have developed from reactive maintenance and preventive maintenance to predictive maintenance. Online detection and fusion calculation of information such as vibration, temperature, and acoustic emission at multiple points of equipment, and real-time fault diagnosis and health status identification are the core and foundation for realizing predictive maintenance of equipment.

[0003] In existing equipment fault diagnosis and health status identification systems, their architectures are mainly centralized or hierarchical from top to bottom. They mainly rely on the main controller to control the cooperation of the equipment diagnosis system, and need to develop customized systems for specific scenarios. There are disadvantages such as complex systems, lack of flexibility and adaptability, over-reliance on the main controller, being prone to getting into local dilemmas, and low optimization efficiency of system cooperation. Summary of the Invention

[0004] The main purpose of the present invention is to provide an adaptive equipment health diagnosis system and diagnosis method based on MAS, mainly to solve the problems of complex systems, lack of flexibility and adaptability, and over-reliance on the main controller in the current diagnosis system. Further, a diagnosis method for applying the system to rotating equipment is provided.

[0005] The object of the present invention can be achieved by adopting the following technical solutions:

[0006] An adaptive equipment health diagnosis system based on MAS,

[0007] The diagnosis system is a single-layer multi-agent system composed of N diagnostic intelligent terminals, where N≥2, and any two of the N diagnostic intelligent terminals are wirelessly or wiredly connected;

[0008] Each diagnostic intelligent terminal includes a signal acquisition module, a signal processing module, a communication module for enabling mutual communication between diagnostic intelligent terminals, and a fusion calculation and analysis module, where the signal acquisition module includes at least one vibration sensor for collecting vibration signals of the equipment;

[0009] Each diagnostic intelligent terminal is internally provided with a main controller function module for commanding and coordinating each intelligent terminal to work collaboratively, so that each diagnostic intelligent terminal has the ability to act as a virtual main controller;

[0010] The diagnostic intelligent terminal further includes a fault source location module, which is used to utilize the cross-correlation principle of the conduction intensity and time sequence of vibration in the physical space of the device, and combine the spatial position of the diagnostic intelligent terminal to perform spatial source location of the fault position;

[0011] Each of the diagnostic intelligent terminals is built-in with an adaptive communication networking module, which is used to perform adaptive networking among multiple diagnostic intelligent terminals, so that the networked diagnostic system can complete the optimal allocation of computing resources and tasks under the command and coordination of the virtual master controller;

[0012] The diagnostic system further includes a health identification and life prediction module, which is used to perform fusion calculation based on the real-time data given by the diagnostic intelligent terminal, and output various health indexes and deterioration trends of the current device, and further combine the deterioration curves of historical monitoring points, the time points, positions and cause elements of fault occurrence and repair for calculation, and finally output the life prediction result;

[0013] The diagnostic system further includes a virtual master controller generation module, which is used to generate a diagnostic intelligent terminal serving as the virtual master controller according to a preset rule or randomly among N diagnostic intelligent terminals, and when the current diagnostic intelligent terminal serving as the virtual master controller cannot work properly due to a fault or exits due to system adjustment, automatically generate a new virtual master controller to perform the system main control work.

[0014] Preferably, the preset rule refers to the rule of resource balance combined with nearby communication.

[0015] Preferably, the communication module of the diagnostic intelligent terminal is also used to enable the terminal to communicate with other devices, so as to realize external data input and external output of health diagnosis data when needed.

[0016] Preferably, each diagnostic intelligent terminal can independently complete signal conversion, fault prediction, and fault cause analysis, and has the ability of self-management and self-regulation, can perform adaptive learning according to different monitoring objects, so as to automatically adjust its own behavior and state, and can continuously learn and modify its own behavior to adapt to the changes of the detected object and environment during the monitoring process.

[0017] Preferably, the signal processing of the signal processing module includes at least performing time-domain and frequency-domain conversion on one or more of the acceleration, velocity, and displacement of the vibration signal, and being able to calculate and output vibration characteristic values.

[0018] Preferably, the signal acquisition module further includes one or more sensors among temperature and acoustic emission sensors.

[0019] Preferably, each of the diagnostic intelligent terminals can independently communicate and interact with other diagnostic intelligent terminals according to their respective intentions, and solve problems in parallel to achieve the purpose of health diagnosis of the equipment.

[0020] Preferably, each of the diagnostic intelligent terminals is communicatively connected to other diagnostic intelligent terminals and directly interacts and cooperates with other diagnostic intelligent terminals, or each of the diagnostic intelligent terminals is only connected to the adjacent diagnostic intelligent terminals and realizes the interaction and cooperation with all diagnostic intelligent terminals through bridge communication.

[0021] A diagnostic method for an adaptive equipment health diagnosis system based on MAS, the diagnostic system being a single-layer multi-agent system composed of N diagnostic intelligent terminals, and the diagnostic system being applied to the fault diagnosis and health assessment of rotating equipment;

[0022] The diagnostic method includes the following steps:

[0023] Step 1: Install diagnostic intelligent terminals 1 to N at the positions to be monitored on the rotating equipment respectively;

[0024] Each of the diagnostic intelligent terminals 1 to N can independently carry out fault diagnosis and health assessment work. Each diagnostic intelligent terminal has a signal acquisition module composed of 1 three-axis acceleration vibration sensor, 1 temperature sensor, 1 acoustic wave sensor and corresponding conversion circuits;

[0025] Each diagnostic intelligent terminal communicates and interacts with other diagnostic intelligent terminals through wireless WIFI;

[0026] Step 2: After installation and power-on, the following and not limited to rated speed, power, number of teeth and other equipment information of the equipment can be input into the diagnostic system one by one on-site;

[0027] Step 3: The diagnostic intelligent terminal starts the diagnostic process, and the temperature and acoustic wave signals collected by the signal acquisition module directly enter the fusion calculation and analysis module with an ESP32 single-chip microcomputer as the core;

[0028] Step 4: The vibration digital signal collected by the signal acquisition module passes through the signal processing module with an STM32 single-chip microcomputer as the core;

[0029] Step 5: Perform Fourier transform calculation on the accelerations in the X, Y, and Z directions of the three-axis acceleration vibration sensor and output the vibration time domain, frequency domain, velocity, velocity RMS, inverse frequency, kurtosis, and impulse information in the X, Y, and Z directions to the fusion calculation and analysis module;

[0030] Step 6: After the initial installation is completed, the fusion calculation and analysis module performs 72 hours of adaptive learning on vibration, temperature, and acoustic wave signals to calculate the time-amplitude curve data of the temperature, acoustic wave, and vibration velocity information at the monitoring positions 1 to N of the diagnostic intelligent terminal under the normal operating state of the rotating equipment and saves it as a reference curve for good working conditions;

[0031] Step 7: After the learning is completed, the system will balance the computing resources according to the self-learning results, and use the diagnostic intelligent terminal with the smallest resource occupancy among the above N diagnostic intelligent terminals as the virtual host. After the virtual host is generated, the system automatically enters the normal health diagnosis and detection stage;

[0032] Step 8.1: When any one of the vibration velocity RMS, temperature, and acoustic wave at a certain monitoring point reaches the shutdown alarm threshold, the corresponding terminal issues a shutdown alarm;

[0033] Step 8.2: When a certain vibration signal is in an abnormal state or in a mutation state, the temperature, acoustic wave, deviation from the good working condition reference curve, kurtosis, and pulse data of this point and adjacent points will be fused for calculation and analysis, and attention alarm, abnormal alarm, or shutdown alarm signals will be issued according to the calculation results;

[0034] Step 9: In Step 8.1 or Step 8.2, the time sequence and intensity cross-correlation of the vibration signals at positions 1 to N are calculated to determine the specific position of the fault source among the 1 to N monitoring points;

[0035] Step 10: The virtual host transmits the deterioration curves of each monitoring point to a mobile phone or other terminal, providing decision-making reference for users' predictive maintenance or targeted preventive maintenance of the generator set.

[0036] Preferably, in Step 8.1 or Step 8.2, the causes of dynamic unbalance, misalignment, or gear damage faults of the equipment failure will be prompted through frequency doubling and inverse frequency analysis.

[0037] The beneficial technical effects of the present invention:

[0038] The diagnostic system provided by the present invention has high self-adaptability and flexibility, can independently complete multi-layer target tasks such as multi-dimensional information perception, data processing, fusion judgment, and AI self-adaptation. The intelligent agent products can automatically handshake and cooperate to complete complex tasks. Users do not need to perform front-end development, and the development complexity and workload of solution providers are greatly reduced. The addition or withdrawal of business nodes, and the increase, decrease, and adjustment of local logic functions do not require rewriting the system software.

[0039] The diagnostic system in the present invention can perform condition monitoring and intelligent diagnosis on equipment and facilities with vibration signals, thereby effectively improving the efficiency and accuracy of equipment fault diagnosis and health status identification, significantly reducing the system development and deployment costs and cycle, and timely meeting the diverse needs of users in different scenarios for equipment diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 Schematic diagram of a steam turbine generator set and a diagnostic system according to an embodiment of the present invention;

[0041] Figure 2 Schematic diagram of wireless networking of diagnostic intelligent terminals according to an embodiment of the present invention;

[0042] Figure 3 Schematic diagram of wired networking of diagnostic intelligent terminals according to an embodiment of the present invention.

[0043] In the figure: 1 - steam turbine, 2 - speed reducer, 3 - generator, 4 - exciter. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0044] To make the technical solutions of the present invention clearer and more definite to those skilled in the art, the present invention will be further described in detail below in conjunction with the embodiments and the accompanying drawings. However, the embodiments of the present invention are not limited thereto.

[0045] As Figures 1 to 3 shown, the adaptive equipment health diagnosis system based on MAS provided in this embodiment

[0046] The diagnostic system is a single-layer multi-agent system composed of 5 diagnostic intelligent terminals.

[0047] Each diagnostic intelligent terminal can independently communicate and interact with other diagnostic intelligent terminals according to their respective intentions, and solve problems in parallel to achieve the purpose of health diagnosis of the equipment;

[0048] Any two of the 5 diagnostic intelligent terminals are wirelessly connected or wired connected;

[0049] Each diagnostic intelligent terminal communicates with other diagnostic intelligent terminals and directly interacts and cooperates with other diagnostic intelligent terminals. As Figure 2 shown, it is suitable for wireless networking, or,

[0050] Each diagnostic intelligent terminal is only connected to the adjacent diagnostic intelligent terminal and realizes interaction and cooperation with all diagnostic intelligent terminals through bridge communication. As Figure 3 shown, it is suitable for wired networking;

[0051] Each diagnostic intelligent terminal includes a signal acquisition module, a signal processing module, a communication module for enabling communication between diagnostic intelligent terminals, and a fusion calculation and analysis module. The signal acquisition module includes at least one vibration sensor, a temperature sensor, and an acoustic emission sensor for collecting vibration signals of the device;

[0052] The signal processing of the signal processing module includes performing time-domain and frequency-domain conversions on at least one or more of the acceleration, velocity, and displacement of the vibration signal, and being able to calculate and output vibration characteristic values such as RMS, kurtosis, and pulse value;

[0053] The communication module of the diagnostic intelligent terminal is also used to enable the terminal to communicate with other devices to achieve external data input and external output of health diagnosis data when needed;

[0054] Each diagnostic intelligent terminal can independently complete functions such as signal conversion, fault prediction, and fault cause analysis, and has the ability of self-management and self-regulation. It can perform adaptive learning according to different monitoring objects, thereby automatically adjusting its own behavior and state, and can continuously learn and modify its behavior to adapt to the changes of the monitored object and the environment during the monitoring process;

[0055] Each diagnostic intelligent terminal is built-in with a main controller function module for commanding and coordinating each intelligent terminal to work collaboratively, so that each diagnostic intelligent terminal has the ability to act as a virtual main controller;

[0056] The diagnostic intelligent terminal also includes a fault source location module, which is used to utilize the mutual correlation principle of the conduction intensity and time sequence of vibration in the physical space of the device, and combine the spatial position of the diagnostic intelligent terminal to perform spatial source location of the fault position;

[0057] Each diagnostic intelligent terminal is built-in with an adaptive communication networking module for adaptively networking multiple diagnostic intelligent terminals, so that the networked diagnostic system mainly solves problems such as task allocation, resource conflict, and knowledge conflict under the command and coordination of the virtual main controller;

[0058] The diagnostic system also includes a health identification and life prediction module for performing fusion calculations based on the real-time data given by the diagnostic intelligent terminal, outputting various health indicators and deterioration trends of the current device, further combining the deterioration curves of historical monitoring points, the time points, positions, and cause elements of fault occurrence and repair, and finally outputting the life prediction result;

[0059] The diagnostic system further includes a virtual master controller generation module, which is used to generate a diagnostic intelligent terminal acting as the virtual master controller according to a preset rule or randomly among N diagnostic intelligent terminals, and when the diagnostic intelligent terminal currently acting as the virtual master controller cannot work properly due to a fault or exits due to system adjustment, a new virtual master controller is automatically generated to perform the system master control work;

[0060] The preset rule refers to the rule of resource balance combined with nearby communication, that is, in the case of wired connection, first ensure that the wired distance between the external communication port of the diagnostic intelligent terminal and the communication port of the external device to be connected is as short as possible, and the virtual master controller is preferably generated among the diagnostic intelligent terminals that need to output the final result, so as to achieve the purpose of resource balance and rapid deployment.

[0061] In the case of wireless connection, first ensure the communication bandwidth and signal quality between the diagnostic intelligent terminal and the external device to be connected. In the case of a small number of terminals and relatively concentrated layout, a random generation mode can be adopted to facilitate rapid deployment.

[0062] As Figure 1 shown, the diagnostic method of the MAS-based adaptive device health diagnostic system provided in this embodiment

[0063] The diagnostic system is a single-layer multi-agent system composed of 5 diagnostic intelligent terminals, and the 5 diagnostic intelligent terminals are the No. ① to ⑤ diagnostic intelligent terminals respectively. This system can perform fault diagnosis and health assessment on the steam turbine generator set.

[0064] The diagnostic method includes the following steps:

[0065] Step 1: Diagnostic intelligent terminal ① and ② are respectively installed at both ends of the steam turbine 1, diagnostic intelligent terminal ③ is installed on the reducer 2, diagnostic intelligent terminal ④ is installed on the generator 3, and diagnostic intelligent terminal ⑤ is installed on the exciter 4;

[0066] Each of the diagnostic intelligent terminals No. ① to ⑤ can independently carry out fault diagnosis and health assessment work. Each diagnostic intelligent terminal has a signal acquisition module composed of 1 three-axis acceleration vibration sensor, 1 temperature sensor, 1 acoustic wave sensor and corresponding conversion circuits;

[0067] Each diagnostic intelligent terminal communicates and interacts with other diagnostic intelligent terminals through wireless WIFI.

[0068] Step 2: After installation and power-on, the relevant device information such as the rated speed, power and number of teeth of the device is input into the diagnostic system one by one according to the prompts of the mobile phone APP supporting this system at the site;

[0069] Step 3: The diagnostic intelligent terminal starts the diagnostic process, and the temperature and acoustic signals collected by the signal acquisition module directly enter the fusion calculation and analysis module with the ESP32 single-chip microcomputer as the core;

[0070] Step 4: The vibration digital signals collected by the signal acquisition module pass through the signal processing module with the STM32 single-chip microcomputer as the core;

[0071] Step 5: Fourier transform calculations are performed on the accelerations in the X, Y, and Z directions of the three-axis acceleration vibration sensor, and information such as the vibration time domain, frequency domain, velocity, velocity RMS, inverse frequency, kurtosis, and impulse in the X, Y, and Z directions are output to the fusion calculation and analysis module;

[0072] Step 6: After the initial installation is completed, the fusion calculation and analysis module performs 72 hours of adaptive learning on signals such as vibration, temperature, and acoustic waves to calculate the time-amplitude curve data of the temperature, acoustic waves, vibration velocity, etc. at the monitoring positions ①~⑤ of the diagnostic intelligent terminal under the normal working state of the steam turbine generator set and saves them as the reference curve for good working conditions;

[0073] Step 7: After the learning is completed, the system will balance the computing resources according to the self-learning results, and use the diagnostic intelligent terminal with the smallest resource occupancy among the above 5 diagnostic intelligent terminals as the virtual host. After the virtual host is generated, the system automatically enters the normal health diagnosis and detection stage;

[0074] Step 8.1: When any value of the vibration velocity RMS, temperature, or acoustic wave at a certain monitoring point reaches the shutdown alarm threshold, the corresponding terminal issues a shutdown alarm;

[0075] Step 8.2: When a certain vibration signal is in an abnormal state or in a mutation state, the temperature, acoustic wave, deviation from the good working condition reference curve, kurtosis, impulse, etc. data of this point and adjacent points will be fused for calculation and analysis, and attention alarm, abnormal alarm, or shutdown alarm signals will be issued according to the calculation results. Further, the causes of equipment faults such as dynamic unbalance, misalignment, and gear damage will be prompted through frequency multiplication and inverse frequency analysis;

[0076] Step 9: In Step 8.1 or Step 8.2, the time series and intensity cross-correlation calculations of the vibration signals at points ①~⑤ are performed to determine the specific position of the fault source among the monitoring points ①~⑤, which helps to quickly determine the accurate location of the fault;

[0077] Step 10: The virtual host transmits the deterioration curves of each monitoring point to the mobile phone or other terminals, providing a decision-making reference for the user's predictive maintenance or targeted preventive maintenance of the generator set.

[0078] In this embodiment, the user can also manually adjust the criteria such as the threshold value preset for diagnosis through the mobile phone APP or other terminals.

[0079] In this embodiment, when each component of the steam turbine is in the adaptive learning stage, in the case that the original moving parts have not passed the running-in period or new application parts are repaired and replaced, each diagnosis will automatically and real-time update the reference curve of each point for good working conditions according to the results of continuous re-learning.

[0080] In summary, in this embodiment, the diagnostic system provided in this embodiment has high self-adaptability and flexibility. The product independently completes multi-dimensional information perception, data processing, fusion judgment, AI self-adaptation and other multi-layer target tasks. The intelligent agent products can automatically handshake and cooperate to complete complex tasks. The user does not need to perform front-end development, and the development complexity and workload of the solution provider are greatly reduced. For the addition or withdrawal of business nodes, the increase, decrease and adjustment of local logic functions do not require rewriting the system software. In view of the characteristics of large uncertainty in the functions and logics of the factory IOT micro-points, it supports taking the manufacturing team as the main body, starting from the pain points at the field bottom layer, and meeting the actual intelligent needs point by point, meeting the differentiated, fragmented and continuously developing system integration needs of the workshop intelligent scenarios. The enterprise team can independently and quickly complete the digital transformation task, complete the system maintenance and iteration work by itself, without customized development, that is, it can perform condition monitoring and intelligent diagnosis on equipment and facilities with vibration signals, thereby effectively improving the efficiency and accuracy of equipment fault diagnosis and health status identification, greatly reducing the system development and deployment costs and cycles, and timely meeting the diverse needs of users in different scenarios for equipment diagnosis.

[0081] The above is only a further embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the scope disclosed by the present invention, according to the technical solution and its concept of the present invention, makes equivalent substitutions or changes, all belong to the protection scope of the present invention.

Claims

1. An MAS-based adaptive device health diagnosis system, characterized in that: the diagnosis system is a single-layer multi-agent system composed of N diagnosis intelligent terminals, where N≥2, and the N diagnosis intelligent terminals are connected by wireless communication or wired connection; each diagnosis intelligent terminal includes a signal acquisition module, a signal processing module, a communication module for enabling mutual communication between diagnosis intelligent terminals, and a fusion calculation and analysis module, where the signal acquisition module includes at least one vibration sensor for acquiring the vibration signal of the device; each diagnosis intelligent terminal is internally provided with a main controller function module for commanding and coordinating each intelligent terminal to work collaboratively, so that each diagnosis intelligent terminal has the ability to act as a virtual main controller; the diagnosis intelligent terminal further includes a fault source location module, which is used to utilize the mutual correlation principle of the conduction intensity and time sequence of vibration in the physical space of the device, and in combination with the spatial position of the diagnosis intelligent terminal, perform spatial source location of the fault position; each diagnosis intelligent terminal is internally provided with an adaptive communication networking module for adaptively networking multiple diagnosis intelligent terminals, so that the networked diagnosis system completes the optimal allocation of computing resources and tasks under the command and coordination of the virtual main controller; the diagnosis system further includes a health identification and life prediction module for performing fusion calculation based on the real-time data given by the diagnosis intelligent terminal, outputting various health indicators and deterioration trends of the current device, and further performing calculations in combination with the deterioration curves of historical monitoring points, the time points, positions and cause elements of fault occurrence and repair, and finally outputting the life prediction result; the diagnosis system further includes a virtual main controller generation module, which is used to generate a diagnosis intelligent terminal acting as a virtual main controller according to a preset rule or randomly among the N diagnosis intelligent terminals, and when the current diagnosis intelligent terminal acting as a virtual main controller cannot work properly due to a fault or exits due to system adjustment, automatically generate a new virtual main controller to perform system main control work.

2. The MAS-based adaptive device health diagnosis system according to claim 1, characterized in that: the preset rule refers to the rule of resource balance combined with nearby communication.

3. The MAS-based adaptive device health diagnosis system according to claim 1, characterized in that: the communication module of the diagnosis intelligent terminal is further used to enable the terminal to communicate with other devices to realize external data input and external output of health diagnosis data when needed.

4. The MAS-based adaptive device health diagnosis system according to claim 1, characterized in that: each diagnosis intelligent terminal can independently complete signal conversion, fault prediction, and fault cause analysis, and has the ability of self-management and self-regulation, can perform adaptive learning according to different monitoring objects, thus automatically adjusting its own behavior and state, and can continuously modify its own behavior through learning to adapt to the changes of the monitored object and environment during the monitoring process.

5. The MAS-based adaptive equipment health diagnosis system according to claim 1, Features: The signal processing of the signal processing module includes performing time domain and frequency domain conversion on at least one or more signals of acceleration, velocity and displacement of the vibration signal, and can calculate and output vibration characteristic values.

6. The MAS-based adaptive equipment health diagnosis system according to claim 1, Features: The signal acquisition module also includes one or more sensors of temperature and acoustic emission sensors.

7. The MAS-based adaptive equipment health diagnosis system according to claim 1, Features: Each of the diagnostic intelligent terminals can autonomously communicate and coordinate with other diagnostic intelligent terminals according to their own intentions, and solve problems in parallel to achieve the purpose of health diagnosis of the equipment.

8. The MAS-based adaptive equipment health diagnosis system according to claim 1, Features: Each of the diagnostic intelligent terminals is communicatively connected with other diagnostic intelligent terminals and directly interacts and cooperates with other diagnostic intelligent terminals, or each of the diagnostic intelligent terminals is only connected to the diagnostic intelligent terminals adjacent to it and interacts and cooperates with all diagnostic intelligent terminals through bridge communication.

9. A diagnostic method for an adaptive equipment health diagnostic system based on MAS, Features: The diagnostic system is a diagnostic system as claimed in any one of claims 1 to 8, and the diagnostic system is applied to fault diagnosis and health assessment of rotating equipment; The diagnostic method comprises the following steps: Step 1: Install the diagnostic intelligent terminals 1 to N at the positions to be monitored on the rotating equipment respectively; Each of the diagnostic intelligent terminals 1 to N can independently carry out fault diagnosis and health assessment work, and each diagnostic intelligent terminal has a signal acquisition module composed of a three-axis acceleration vibration sensor, a temperature sensor, a sound wave sensor and corresponding conversion circuits; Each diagnostic intelligent terminal communicates and coordinates with other diagnostic intelligent terminals via wireless WIFI; Step 2: After the installation is completed and powered on, the equipment information including but not limited to the rated speed, power and number of teeth can be input into the diagnostic system one by one on site; Step 3: The diagnostic intelligent terminal starts the diagnostic process. The temperature and sound wave signals collected by the signal acquisition module directly enter the fusion calculation and analysis module with the ESP32 microcontroller as the core; Step 4: The vibration digital signal collected by the signal acquisition module passes through a signal processing module with the STM32 single-chip microcomputer as the core; Step 5, perform Fourier transform calculation on the acceleration in the X, Y, and Z directions of the three-axis acceleration vibration sensor and output the vibration time domain, frequency domain, velocity and velocity RMS, inverse frequency, kurtosis, and pulse information in the X, Y, and Z directions to the fusion calculation and analysis module; Step 6: After the initial installation is completed, the fusion calculation and analysis module performs 72-hour adaptive learning on vibration, temperature, and acoustic wave signals to calculate the time-amplitude curve data of temperature, acoustic wave, and vibration speed information at the monitoring positions 1 to N of the diagnostic intelligent terminal under the normal working state of the rotating equipment and save it as a reference curve for good working conditions; Step 7: After the learning is completed, the system will balance the computing resources according to the self-learning results, and use the diagnostic intelligent terminal with the smallest resource occupancy among the above N diagnostic intelligent terminals as the virtual host. After the virtual host is generated, the system automatically enters the normal health diagnosis and detection stage; Step 8.1: When any one of the vibration speed RMS, temperature, and acoustic wave at a certain monitoring point reaches the shutdown alarm threshold, the corresponding terminal issues a shutdown alarm; Step 8.2: When a certain vibration signal is in an abnormal state or a mutation state, the temperature, acoustic wave, deviation degree from the good working condition reference curve, kurtosis, and pulse data at this point and adjacent points will be fused for calculation and analysis, and a attention alarm, abnormal alarm, or shutdown alarm signal will be issued according to the calculation results; Step 9: In Step 8.1 or Step 8.2, the time sequence and intensity cross-correlation of the vibration signals at positions 1 to N are calculated to determine the specific position of the fault source among the 1 to N monitoring points; Step 10: The virtual host transmits the deterioration curves of each monitoring point to a mobile phone or other terminal, providing a decision-making reference for the user's predictive maintenance or targeted preventive maintenance of the generator set.

10. The diagnostic method of the MAS-based adaptive equipment health diagnosis system according to claim 9, characterized in that: In Step 8.1 or Step 8.2, the reasons for dynamic unbalance of equipment faults, misalignment, or gear damage faults will be prompted through octave and inverse frequency analysis.

Images