A method and device for predictive maintenance of equipment based on joint learning

Abstract

The present invention discloses a method, device, readable medium and electronic device for predictive maintenance of equipment based on joint learning. The method comprises: obtaining a knowledge graph of a target equipment; sending the knowledge graph of the target equipment to a joint learning center, so that the joint learning center determines an integrated knowledge graph based on the knowledge graphs of the target equipment sent by multiple target data participants; determining a fault prediction function based on the received integrated knowledge graph sent by the joint learning center; and obtaining the probability of failure of the target equipment based on the fault prediction function. The technical solution provided by the present invention obtains the integrated knowledge graph through the joint learning center, and the probability of failure of the target equipment obtained using the integrated knowledge graph has a higher accuracy.

Description

Technical Field

[0001] The present invention relates to the field of energy, and in particular to a method and device for predictive maintenance of equipment based on joint learning. Background Art

[0002] There are a large number of devices in the integrated energy system, and each device is usually not independent. If any device has a problem, multiple devices may not work properly. Therefore, the normal operation of the equipment is a prerequisite for the efficient operation of the integrated energy system. In order to ensure the normal operation of the equipment, it is usually necessary to perform predictive maintenance on the equipment, that is, before the equipment fails, the operating status of the equipment is judged to predict the possible failure of the equipment. At present, when performing predictive maintenance on equipment, it is often necessary to obtain a large amount of sample data and build a physical model. However, when using the physical model for prediction, the calculation is complex and time-consuming. For complex equipment, it may not be possible to obtain the physical model of the equipment. Therefore, it is crucial to determine a new method for predictive maintenance of equipment. Summary of the invention

[0003] The present invention provides a method, device, readable medium and electronic device for predictive maintenance of equipment based on joint learning. An integrated knowledge graph is obtained through a joint learning center. The probability of failure of the target equipment obtained using the integrated knowledge graph has higher accuracy.

[0004] In a first aspect, the present invention provides a method for predictive maintenance of equipment based on joint learning, comprising:

[0005] Obtain the knowledge graph of the target device;

[0006] Sending the knowledge graph of the target device to a joint learning center, so that the joint learning center determines an integrated knowledge graph based on the knowledge graphs of the target devices sent by multiple target data participants;

[0007] Determine a fault prediction function based on the received integrated knowledge graph sent by the joint learning center;

[0008] Based on the fault prediction function, the probability of a target device failing is obtained.

[0009] Preferably,

[0010] The method further comprises:

[0011] Determine a descriptive word of the target device, wherein the descriptive word includes at least one of a device model of the target device, a task type of the target device, and a location of the target device;

[0012] Then, sending the knowledge graph of the target device to the joint learning center so that the joint learning center determines an integrated knowledge graph based on the knowledge graphs of the target devices sent by multiple target data participants, includes:

[0013] The knowledge graph carrying the descriptive words is sent to the joint learning center, so that the joint learning center determines the target data participant based on the descriptive words, and determines the integrated knowledge graph based on the knowledge graph sent by the target data participant.

[0014] Preferably,

[0015] The step of obtaining the knowledge graph of the target device includes:

[0016] Determine local fault data of target equipment;

[0017] Obtain the collected fault data of the device type corresponding to the target device;

[0018] Based on the local fault data and the collected fault data, a knowledge graph of the target device is determined.

[0019] Preferably,

[0020] The method further comprises:

[0021] Determine the knowledge graph of the associated devices of the target device;

[0022] Sending the knowledge graph of the associated device to a joint learning center, so that the joint learning center determines the integrated system knowledge graph based on the knowledge graphs of the target device and the knowledge graphs of the associated device sent by multiple target data participants;

[0023] Based on the received integrated system knowledge graph sent by the joint learning center, the probability of a system failure is determined, and the system includes a target device and associated devices.

[0024] Preferably,

[0025] The method further comprises:

[0026] Determine the predicted failure type of the target device;

[0027] Based on the integrated knowledge graph, determining an operation and maintenance plan corresponding to the predicted fault type;

[0028] If the target component exists in the operation and maintenance plan, determine the current inventory status of the target component;

[0029] If the current inventory status meets the preset conditions, a prompt message is sent.

[0030] Preferably,

[0031] The method further comprises:

[0032] Determine the current operation information of the target device;

[0033] Based on the current operation information and the fault prediction function, a fault time of the target device is determined.

[0034] Preferably,

[0035] The determining of the fault prediction function based on the received integrated knowledge graph sent by the joint learning center includes:

[0036] Determine a fault tree based on the received integrated knowledge graph sent by the joint learning center;

[0037] Based on the fault tree, a fault prediction function is determined.

[0038] In a second aspect, the present invention provides a device for predictive maintenance of equipment based on joint learning, comprising:

[0039] A graph acquisition module, used to acquire the knowledge graph of the target device;

[0040] A graph sending module, used to send the knowledge graph of the target device to the joint learning center, so that the joint learning center determines an integrated knowledge graph based on the knowledge graphs of the target devices sent by multiple target data participants;

[0041] A function determination module, configured to determine a fault prediction function based on the received integrated knowledge graph sent by the joint learning center;

[0042] The probability acquisition module is used to acquire the probability of a target device failing based on the failure prediction function.

[0043] In a third aspect, the present invention provides a readable medium, comprising execution instructions. When a processor of an electronic device executes the execution instructions, the electronic device executes any method described in the first aspect.

[0044] In a fourth aspect, the present invention provides an electronic device, comprising a processor and a memory storing execution instructions, wherein when the processor executes the execution instructions stored in the memory, the processor executes any method described in the first aspect.

[0045] The present invention provides a method, device, readable medium and electronic device for predictive maintenance of equipment based on joint learning. The method obtains the knowledge graph of the target equipment and sends the knowledge graph to the joint learning center. The joint learning center receives the knowledge graph and determines multiple target data participants. The joint learning center determines the integrated knowledge graph using the knowledge graphs sent by the multiple target data participants, and then sends the integrated knowledge graph to each target data participant. After receiving the integrated knowledge graph sent by the joint learning center, the target data participant determines a fault prediction function, and then obtains the probability of failure of the target equipment according to the fault prediction function. The technical solution provided by the present invention integrates the knowledge graphs of each target data participant through the joint learning center to obtain the integrated knowledge graph. The probability of failure of the target equipment obtained using the integrated knowledge graph is more accurate, and the technical solution does not need to obtain the physical model of the target equipment. Even complex target equipment can still use the technical solution provided by the present invention for predictive maintenance. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0047] Figure 1 A schematic flow chart of a first method for predictive maintenance of equipment based on joint learning provided in an embodiment of the present invention;

[0048] Figure 2 A schematic flow chart of a second method for predictive maintenance of equipment based on joint learning provided in an embodiment of the present invention;

[0049] Figure 3 A schematic flow chart of a third method for predictive maintenance of equipment based on joint learning provided in an embodiment of the present invention;

[0050] Figure 4 A schematic flow chart of a fourth method for predictive maintenance of equipment based on joint learning provided in an embodiment of the present invention;

[0051] Figure 5 A schematic flow chart of a fifth method for predictive maintenance of equipment based on joint learning provided in an embodiment of the present invention;

[0052] Figure 6 A schematic diagram of the structure of a device predictive maintenance device based on joint learning provided in an embodiment of the present invention;

[0053] Figure 7 The present invention is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0054] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0055] like Figure 1 As shown, an embodiment of the present invention provides a method for predictive maintenance of equipment based on joint learning, the method comprising:

[0056] Step 11, obtaining the knowledge graph of the target device;

[0057] Step 12, sending the knowledge graph of the target device to the joint learning center, so that the joint learning center determines an integrated knowledge graph based on the knowledge graphs of the target devices sent by multiple target data participants;

[0058] Step 13, determining a fault prediction function based on the received integrated knowledge graph sent by the joint learning center;

[0059] Step 14: Based on the fault prediction function, obtain the probability of a target device failing.

[0060] In the above embodiment, by acquiring the knowledge graph of the target device, the logical relationship between the faults, anomalies and phenomena of the device will be clarified in the knowledge graph, and then the knowledge graph is sent to the joint learning center. At this time, the current user is a data participant of the joint learning center. The joint learning center receives the knowledge graph and determines multiple target data participants (including the current user). The joint learning center uses the knowledge graphs sent by multiple target data participants to determine the integrated knowledge graph, and then the joint learning center sends the integrated knowledge graph to each target data participant. After receiving the integrated knowledge graph sent by the joint learning center, the target data participant determines the fault prediction function. When determining the fault prediction function, the fault tree can be determined based on the integrated knowledge graph sent by the joint learning center; based on the fault tree, the fault prediction function is determined according to the Monte Carlo method. Then, according to the fault prediction function, the probability of the target device failing is obtained. The technical solution provided in this embodiment integrates the knowledge graphs of various target data participants through a joint learning center to obtain an integrated knowledge graph. The probability of failure of the target equipment obtained using the integrated knowledge graph is more accurate, and there is no need to obtain the physical model of the target equipment. Even for complex target equipment, predictive maintenance can still be performed using the technical solution provided in this embodiment.

[0061] Specifically, when using the fault prediction function to determine the probability of a target device failing, you can use a simulated device operation method to determine it, such as simulating the operation 1,000 times to obtain the probability of the target device failing over time. Therefore, when determining the probability of a target device failing, it is often necessary to determine the time of the failure, and the time of the failure is related to the current operation information of the target device. Based on the current operation information, determine the operating age of the device, and determine the failure time of the target device based on the operating age of the device, so as to ensure that relevant staff can perform timely predictive maintenance.

[0062] like Figure 2 As shown, in one embodiment of the present invention, the method further includes:

[0063] Step 15, determining a descriptive word of the target device, wherein the descriptive word includes at least one of a device model of the target device, a task type of the target device, and a location of the target device;

[0064] Then, the step 12 sends the knowledge graph of the target device to the joint learning center, so that the joint learning center determines the integrated knowledge graph based on the knowledge graphs of the target devices sent by multiple target data participants, including:

[0065] Step 121, sending the knowledge graph carrying the descriptive words to the joint learning center, so that the joint learning center determines the target data participant based on the descriptive words, and determines the integrated knowledge graph based on the knowledge graph sent by the target data participant.

[0066] In the above embodiment, in order to make the integrated knowledge graph obtained by the joint learning center more accurate, when the data participant sends the knowledge graph to the joint learning center, the sent knowledge graph carries the descriptive words about the target device, so that the joint learning center can determine the target data participant based on the descriptive words, so that the knowledge graphs between the determined target data participants have similarity. The descriptive words include at least one of the device model of the target device, the task type of the target device, and the location of the target device. Different device models may correspond to different faults; even when the same model of equipment performs different tasks, different faults may occur; the location of the target device is different, that is, the production environment of the target device is different, and different production environments have different effects on the device, resulting in different faults in the device. Therefore, in order to ensure the accuracy of the integrated knowledge graph, data participants with similar descriptive words can be selected as target data participants. Therefore, when the joint learning center selects the target data participant, the similarity between the description words of the selected data participant and the current data participant corresponding to the current user is determined, and the selected data participant whose similarity exceeds the preset threshold is selected together with the current data participant as the target data participant.

[0067] like Figure 3 As shown, in one embodiment of the present invention, the step 11 of acquiring the knowledge graph of the target device includes:

[0068] Step 111, determining local fault data of the target device;

[0069] Step 112, obtaining collected fault data of a device type corresponding to the target device;

[0070] Step 113: determine the knowledge graph of the target device based on the local fault data and the collected fault data.

[0071] In the above embodiment, when determining the knowledge graph of the target device, not only the local fault data of the local target device set in the integrated energy system is used, but also the collected fault data is used. The collected fault data can be the historical fault data of the target device collected by the equipment operation manufacturer (such as the equipment's return rate, maintenance time, replacement time, etc.), and the knowledge graph determined by using the local fault data and the collected fault data is more accurate. In one possible situation, the target device is a newly introduced device in the integrated energy system. At this time, there is no local fault data for the target device. At this time, it is particularly important to use the collected fault data collected from the equipment operation manufacturer to determine the knowledge graph of the target device. Of course, if the amount of local fault data of the target device is relatively abundant, the knowledge graph of the target device can be determined using only the local fault data of the target device.

[0072] In one possible implementation, for devices whose executed task types have a significant impact on device failures, devices of different models that execute different task types are determined as different target devices. When determining the knowledge graph, historical failure data of devices with the same device model and the same task type can be selected as local failure data, and the knowledge graph of the target device can be determined using the local failure data. At this time, when the knowledge graph of the target device is sent to the joint learning center, the accuracy of the description word similarity can be determined. For example, in the joint learning center, data participants with the same device type and task type as the current user are determined as target data participants, so that the joint learning center can be used to determine the integrated knowledge graph when the device type executes the task type, thereby ensuring the accuracy of the determined integrated knowledge graph.

[0073] In another possible implementation, the determined target device may include all devices with the same name set up in the integrated energy system. For example, if the target device is determined to be a condenser, the historical fault data of condensers of all models performing all tasks in the integrated energy system can be used as local fault data of the target device, and the knowledge graph can be jointly determined with the collected fault data about the condenser determined at the equipment operating manufacturer. When the knowledge graph of the target device is sent to the joint learning center, the accuracy of the description word similarity can be determined as the description word is a condenser. At this time, it is ensured that the determined integrated knowledge graph has a sufficiently wide dimension to determine any possible faults of the condenser.

[0074] like Figure 4 As shown, in one embodiment of the present invention, the method further includes:

[0075] Step 17, determining the knowledge graph of the associated devices of the target device;

[0076] Step 18, sending the knowledge graph of the associated device to the joint learning center, so that the joint learning center determines the integrated system knowledge graph based on the knowledge graphs of the target device and the knowledge graphs of the associated device sent by multiple target data participants;

[0077] Step 19, based on the received integrated system knowledge graph sent by the joint learning center, determine the probability of system failure, wherein the system includes a target device and associated devices.

[0078] In the above embodiment, it is necessary not only to determine the knowledge graph of the target device, but also to determine the knowledge graph of the associated device that has an associated relationship with the target device, and send both the knowledge graph of the target device and the knowledge graph of the associated device to the joint learning center, so that the joint learning center can determine the integrated system knowledge graph based on the knowledge graph of the target device and the knowledge graph of the associated device sent by multiple target data participants. After obtaining the integrated system knowledge graph, the joint learning center sends the integrated system knowledge graph to the target data participants. After each target data participant receives the integrated system knowledge graph sent by the joint learning center, it determines the fault prediction function corresponding to the system, and then determines the probability of system failure based on the fault prediction function corresponding to the system. The use of the integrated system knowledge graph can more accurately determine the probability of system failure, ensuring that the system can operate normally and efficiently.

[0079] In one embodiment, after obtaining the knowledge graph of the target device and the knowledge graph of the associated device, the knowledge graph of the target device and the knowledge graph of the associated device are fused to determine the fused knowledge graph, and the fused knowledge graph is sent to the joint learning center so that the joint learning center determines the integrated system knowledge graph based on the fused knowledge graphs sent by multiple target data participants, the knowledge graph of the target device or the knowledge graph of the associated device, and then sends the integrated system knowledge graph to the target data participant in the joint learning so that the target data participant determines the probability of system failure based on the integrated system knowledge graph.

[0080] like Figure 5 As shown, in one embodiment of the present invention, after obtaining the probability of a target device failing based on the fault prediction function, the method further includes:

[0081] Step 20, determining the predicted fault type of the target device;

[0082] Step 21: determining an operation and maintenance solution corresponding to the predicted fault type based on the integrated knowledge graph;

[0083] Step 22: if the target component exists in the operation and maintenance plan, determine the current inventory status of the target component;

[0084] Step 23: If the current inventory status meets the preset conditions, a prompt message is sent.

[0085] In the above embodiment, the probability of failure of the target device can be obtained by using the fault prediction function, and then the predicted fault type of the target device can be further determined. When determining the knowledge graph of the target device, the knowledge graph carries the operation and maintenance plan corresponding to the target device failure. Therefore, the integrated knowledge graph determined by the joint learning center also contains the operation and maintenance plan corresponding to the target device failure. After determining the predicted fault type of the target device, the operation and maintenance plan corresponding to the predicted fault type can be determined according to the integrated knowledge graph, and it is further determined whether there is a target accessory that needs to be replaced in the operation and maintenance plan. If the target accessory does not exist in the operation and maintenance plan, the operation and maintenance plan can be sent to the relevant staff. If the target accessory exists in the operation and maintenance plan, the current inventory status of the target accessory is determined. When the current inventory status of the target accessory meets the preset conditions, that is, when the current inventory status is less than the usage quantity of the target accessory in the operation and maintenance plan, a prompt message is sent to the relevant staff to enable the relevant staff to prepare the relevant number of target accessories, thereby determining that the target device can be efficiently operated and maintained.

[0086] Based on the same inventive concept as the above method, Figure 6 As shown, an embodiment of the present invention provides a device for predictive maintenance of equipment based on joint learning, including:

[0087] A graph acquisition module 61, used to acquire a knowledge graph of a target device;

[0088] A graph sending module 62, configured to send the knowledge graph of the target device to a joint learning center, so that the joint learning center determines an integrated knowledge graph based on the knowledge graphs of the target devices sent by multiple target data participants;

[0089] A function determination module 63, configured to determine a fault prediction function based on the received integrated knowledge graph sent by the joint learning center;

[0090] The probability acquisition module 64 is used to acquire the probability of a target device failing based on the failure prediction function.

[0091] For the convenience of description, the above device embodiments are described by dividing the functions into various units or modules respectively. When implementing the present invention, the functions of each unit or module can be implemented in the same or multiple software and / or hardware.

[0092] Figure 7It is a structural diagram of an electronic device provided by an embodiment of the present invention. At the hardware level, the electronic device includes a processor 701 and a memory 702 storing execution instructions, and optionally also includes an internal bus 703 and a network interface 704. Among them, the memory 702 may include a memory 7021, such as a high-speed random access memory (Random-Access Memory, RAM), and may also include a non-volatile memory 7022 (non-volatile memory), such as at least one disk storage, etc.; the processor 701, the network interface 704 and the memory 702 can be interconnected through an internal bus 703, and the internal bus 703 can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc.; the internal bus 703 can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 7 Only one bidirectional arrow is used in the figure, but it does not mean that there is only one bus or one type of bus. Of course, the electronic device may also include hardware required for other services. When the processor 701 executes the execution instruction stored in the memory 702, the processor 701 executes the method in any embodiment of the present invention, and is at least used to execute the following steps: Figures 1 to 5 The method shown.

[0093] In a possible implementation, the processor reads the corresponding execution instructions from the non-volatile memory into the memory and then runs them, and can also obtain the corresponding execution instructions from other devices to form a device predictive maintenance apparatus based on joint learning at the logical level. The processor executes the execution instructions stored in the memory to implement a device predictive maintenance method based on joint learning provided in any embodiment of the present invention through the executed execution instructions.

[0094] The processor may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in the processor or instructions in the form of software. The above processor may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The disclosed methods, steps and logic block diagrams in the embodiments of the present invention can be implemented or executed. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0095] The embodiment of the present invention further provides a computer-readable storage medium, including an execution instruction. When a processor of an electronic device executes the execution instruction, the processor executes the method provided in any embodiment of the present invention. The electronic device may be specifically as follows: Figure 7 The electronic device shown; the execution instruction is a computer program corresponding to a device for predictive maintenance of equipment based on joint learning.

[0096] Those skilled in the art will appreciate that the embodiments of the present invention may be provided as methods or computer program products. Therefore, the present invention may be implemented in the form of a complete hardware embodiment, a complete software embodiment, or a combination of software and hardware.

[0097] The various embodiments of the present invention are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0098] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or boiler including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or boiler. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or boiler including the elements.

[0099] The above description is only an embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent substitution, improvement, etc. made within the spirit and principle of the present invention should be included in the scope of the claims of the present invention.

Claims

1. A method for predictive maintenance of equipment based on joint learning, characterized in that: include: Obtain the knowledge graph of the target device; Sending the knowledge graph of the target device to a joint learning center, so that the joint learning center determines an integrated knowledge graph based on the knowledge graphs of the target devices sent by multiple target data participants; Determine a fault prediction function based on the received integrated knowledge graph sent by the joint learning center; Based on the fault prediction function, obtaining the probability of a target device failing; The method further comprises: Determining a description of a target device, wherein the description includes at least one of a device model of the target device, a task type of the target device, and a location of the target device; sending the knowledge graph of the target device to a joint learning center so that the joint learning center determines an integrated knowledge graph based on the knowledge graphs of the target devices sent by multiple target data participants, including: The knowledge graph carrying the descriptive words is sent to the joint learning center, so that the joint learning center determines the target data participant based on the descriptive words, and determines the integrated knowledge graph based on the knowledge graph sent by the target data participant.

2. The equipment predictive maintenance method based on joint learning according to claim 1 is characterized in that: The step of obtaining the knowledge graph of the target device includes: Determine local fault data of target equipment; Obtain the collected fault data of the device type corresponding to the target device; Based on the local fault data and the collected fault data, a knowledge graph of the target device is determined.

3. The equipment predictive maintenance method based on joint learning according to claim 1 is characterized in that: The method further comprises: Determine the knowledge graph of the associated devices of the target device; Sending the knowledge graph of the associated device to a joint learning center, so that the joint learning center determines the integrated system knowledge graph based on the knowledge graphs of the target device and the knowledge graphs of the associated device sent by multiple target data participants; Based on the received integrated system knowledge graph sent by the joint learning center, the probability of a system failure is determined, and the system includes a target device and associated devices.

4. The equipment predictive maintenance method based on joint learning according to claim 1 is characterized in that: After obtaining the probability of a target device failing based on the fault prediction function, the method further includes: Determine the predicted failure type of the target device; Based on the integrated knowledge graph, determining an operation and maintenance plan corresponding to the predicted fault type; If the target component exists in the operation and maintenance plan, determine the current inventory status of the target component; If the current inventory status meets the preset conditions, a prompt message is sent.

5. The equipment predictive maintenance method based on joint learning according to claim 1 is characterized in that: The method further comprises: Determine the current operation information of the target device; Based on the current operation information and the fault prediction function, a fault time of the target device is determined.

6. The equipment predictive maintenance method based on joint learning according to any one of claims 1 to 5, characterized in that: The determining of the fault prediction function based on the received integrated knowledge graph sent by the joint learning center includes: Determine a fault tree based on the received integrated knowledge graph sent by the joint learning center; Based on the fault tree, a fault prediction function is determined.

7. A device for predictive maintenance of equipment based on joint learning, characterized in that: include: A graph acquisition module, used to acquire the knowledge graph of the target device; A graph sending module, used to send the knowledge graph of the target device to the joint learning center, so that the joint learning center determines an integrated knowledge graph based on the knowledge graphs of the target devices sent by multiple target data participants; A function determination module, configured to determine a fault prediction function based on the received integrated knowledge graph sent by the joint learning center; A probability acquisition module, used to acquire the probability of a target device failing based on the failure prediction function; The graph sending module is also specifically used to: determine a descriptive word of the target device, the descriptive word including at least one of a device model of the target device, a task type of the target device, and a location of the target device; send the knowledge graph carrying the descriptive word to the joint learning center, so that the joint learning center determines the target data participant based on the descriptive word, and determines the integrated knowledge graph based on the knowledge graph sent by the target data participant.

8. A readable medium comprising an execution instruction. When a processor of an electronic device executes the execution instruction, the electronic device executes the method according to any one of claims 1 to 6.

9. An electronic device, comprising a processor and a memory storing execution instructions, wherein when the processor executes the execution instructions stored in the memory, the processor executes the method according to any one of claims 1 to 6.

Images