Equipment failure prediction method, device, readable medium and electronic device

Abstract

The present invention discloses a method, device, readable medium and electronic device for predicting equipment failure. The method includes: determining the characteristic information of the target equipment and the detection point data information corresponding to the target equipment; determining the characteristic data probability distribution model of the target equipment and the probability distribution model of non-shared data based on the characteristic information of the target equipment and the detection point data information corresponding to the target equipment; determining the weight of non-shared data according to the characteristic data probability distribution model and the probability distribution model of the detection point data; establishing a federated learning model according to the non-shared data, the weight of the non-shared data and the equipment failure label corresponding to the non-shared data; predicting the equipment failure of the target equipment according to the federated learning model. The technical solution provided by the present invention can migrate the non-shared data to the target equipment, establish the relationship between the characteristic data of the target equipment and the equipment failure, and ensure data security without sharing the characteristic data between the equipment.

Description

Technical Field

[0001] The present invention relates to the field of energy technology, and in particular to a method, a device, a readable medium and an electronic device for predicting equipment failure. Background Art

[0002] Once an industrial equipment in an intelligent manufacturing system fails, it will affect the normal operation of the intelligent manufacturing system and cause very serious consequences. Therefore, it is usually necessary to predict equipment failures of industrial equipment.

[0003] At present, when making predictions for a certain industrial equipment, since the operating data of the industrial equipment has no labeled data, it is necessary to collect the historical operating data of several other industrial equipment and the label data corresponding to the historical operating data, and use machine learning methods to establish a mapping relationship between these historical operating data and equipment failure predictions, and obtain an equipment failure prediction model. The equipment failure prediction model is used to realize the equipment failure prediction of the industrial equipment.

[0004] However, in order to learn an effective model, the above technical solution may need to share historical operation data of industrial equipment and label data corresponding to the historical operation data when predicting equipment failures of multiple industrial equipment, resulting in low data security between industrial equipment. Summary of the invention

[0005] In view of the above-mentioned technical problems existing in the prior art, the present invention provides a device failure prediction method, device, computer-readable storage medium and electronic device, which can migrate non-shared data to the target device, establish the relationship between the characteristic data of the target device and the device failure, and ensure data security without the need for sharing the characteristic data between devices.

[0006] In a first aspect, the present invention provides a method for predicting equipment failure, comprising:

[0007] Determine characteristic information of the target device and detection point data information corresponding to the target device;

[0008] Determine a probability distribution model of the characteristic data of the target device and a probability distribution model of the detection point data with non-shared data based on the characteristic information of the target device and the detection point data information corresponding to the target device;

[0009] Determining the weight of the non-shared data according to the probability distribution model of the feature data and the probability distribution model of the detection point data;

[0010] Establishing a federated learning model according to the non-shared data, the weight of the non-shared data, and the device fault label corresponding to the non-shared data;

[0011] The device failure prediction of the target device is performed according to the federated learning model.

[0012] In a second aspect, the present invention provides a device for predicting equipment failure, comprising:

[0013] An information determination module, used to determine characteristic information of a target device and detection point data information corresponding to the target device;

[0014] A probability model determination module, used to determine a probability distribution model of characteristic data of the target device and a probability distribution model of detection point data with non-shared data based on the characteristic information of the target device and the detection point data information corresponding to the target device;

[0015] A weight determination module, used to determine the weight of the non-shared data according to the probability distribution model of the feature data and the probability distribution model of the detection point data;

[0016] A model building module, used to build a federated learning model according to the non-shared data, the weight of the non-shared data and the device fault label corresponding to the non-shared data;

[0017] A prediction module is used to predict device failure of the target device according to the federated learning model.

[0018] In a third aspect, the present invention provides a computer-readable storage medium comprising execution instructions. When a processor of an electronic device executes the execution instructions, the processor executes any method described in the first aspect or any method described in the second aspect.

[0019] 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 or any method described in the second aspect.

[0020] The present invention provides a device failure prediction method, apparatus, computer-readable storage medium and electronic device, which determines the characteristic information of a target device and the detection point data information corresponding to the target device, and then determines the characteristic data probability distribution model of the target device and the probability distribution model of the detection point data with non-shared data based on the characteristic information of the target device and the detection point data information corresponding to the target device, and then determines the weight of the non-shared data according to the characteristic data probability distribution model and the probability distribution model of the detection point data, and then establishes a federated learning model according to the non-shared data, the weight of the non-shared data and the device failure label corresponding to the non-shared data, and finally predicts the device failure of the target device according to the federated learning model. The technical solution provided by the present invention establishes the relationship between the characteristic data of the target device and the device failure by migrating the non-shared data to the target device, and does not need to share the characteristic data between devices, thereby ensuring data security.

[0021] The further effects of the above-mentioned non-conventional preferred manner will be described below in conjunction with specific embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the embodiments of the present invention or the existing technical solutions, 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 recorded in the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0023] Figure 1 A schematic diagram of a flow chart of a method for predicting equipment failures provided by an embodiment of the present invention;

[0024] Figure 2 A schematic flow chart of another device failure prediction method provided by an embodiment of the present invention;

[0025] Figure 3 A schematic diagram of the structure of a device for predicting equipment failures provided by an embodiment of the present invention;

[0026] Figure 4 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0027] 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.

[0028] like Figure 1 As shown, an embodiment of the present invention provides a method for predicting equipment failures, comprising the following steps:

[0029] Step 101: Determine feature information of a target device and detection point data information corresponding to the target device.

[0030] Specifically, the target device is any industrial equipment of the intelligent manufacturing system, and the embodiment of the present invention does not specifically limit this. Preferably, the target device is an energy device, such as a gas boiler, an internal combustion engine, a steam turbine, a cogeneration device, a photovoltaic device, etc.

[0031] Specifically, the characteristic information of the target device refers to the attributes or functions of the characteristic data of the target device.

[0032] Specifically, the detection point data information refers to the attributes or functions of the detection point data.

[0033] It should be noted that the detection point data includes non-shared data.

[0034] Step 102: Based on the feature information of the target device and the detection point data information corresponding to the target device, determine a probability distribution model of the feature data of the target device and a probability distribution model of the detection point data with non-shared data.

[0035] Specifically, there are multiple feature data of the target device, and each feature data includes several features corresponding to the feature values ​​of the target device. Among them, several features are influencing factors of equipment failure, which need to be determined in combination with the actual scenario. For example, if the target device is a gas boiler, the several features include but are not limited to gas flow, gas temperature, exhaust temperature, flue gas flow, gas pressure, start and stop status, flue gas humidity and flue gas pressure, etc.

[0036] Specifically, the feature data obeys a continuous probability distribution, and therefore, the feature data probability distribution model may be a normal distribution model or an exponential distribution model, which is not limited in the embodiments of the present invention. Preferably, the feature data probability distribution model is a mixed Gaussian model formed by mixing multiple normal distribution models, wherein the normal distribution model is also called a Gaussian distribution model.

[0037] Specifically, the probability distribution model of the detection point data is the same as the probability distribution model of the feature data. Preferably, it can be a mixed Gaussian model. Of course, it can also be other probability distribution models. The specific needs to be determined in combination with actual conditions and are not specifically limited here.

[0038] In one embodiment of the present invention, the determining of the feature data probability distribution model of the target device and the probability distribution model of the detection point data with non-shared data based on the feature information of the target device and the detection point data information corresponding to the target device comprises:

[0039] Determining the characteristic data of the target device and the detection point data with non-shared data based on the characteristic information of the target device and the detection point data information corresponding to the target device;

[0040] Calculate the data distribution of the characteristic data according to the characteristic data parameter model, and determine the characteristic data parameter model with determined parameters as the characteristic data probability distribution model;

[0041] The data distribution of the detection point data is calculated according to the detection point data parameter model, and the detection point data parameter model with determined parameters is determined as the probability distribution model of the detection point data.

[0042] Specifically, the characteristic data parameter model can be a Gaussian model or a mixed Gaussian model, preferably a mixed Gaussian model. The detection point data parameter model is the same as the characteristic data parameter model, and the embodiment of the present invention will not elaborate on this.

[0043] Specifically, the feature data probability distribution model determines the model parameters by substituting each feature data into a mixed Gaussian model for calculation, wherein the calculation method may be an EM algorithm, and the mixed Gaussian model with determined model parameters is determined as the feature data probability distribution model.

[0044] The probability distribution model of the detection point data is determined by substituting multiple detection point data into a mixed Gaussian model for calculation to determine model parameters, wherein the calculation method may be an EM algorithm, and the mixed Gaussian model with determined model parameters is determined as the probability distribution model of the detection point data.

[0045] Step 103: Determine the weight of the non-shared data according to the probability distribution model of the feature data and the probability distribution model of the detection point data.

[0046] In this embodiment, the non-shared data of the detection point is migrated to the target device through the weight of the non-shared data. In other words, a data association is established between the detection point and the target device. The detection point will not directly obtain the characteristic data of the target device, that is, there is no data sharing between the target device and the detection point, thereby ensuring data security.

[0047] In one embodiment of the present invention, determining the weight of the non-shared data according to the probability distribution model of the feature data and the probability distribution model of the detection point data includes:

[0048] Determining a target device distribution probability of the non-shared data according to the characteristic data probability distribution model;

[0049] Determining the detection point distribution probability of the non-shared data according to the probability distribution model of the detection point data;

[0050] A ratio of the target device distribution probability of the non-shared data to the detection point distribution probability of the non-shared data is determined as the weight of the non-shared data.

[0051] In this embodiment, the weight of the non-shared data is determined by the ratio of the target device distribution probability of the non-shared data and the detection point distribution probability of the non-shared data, so that the non-shared data on the detection point is migrated to the target device. At the same time, there is no need to share the characteristic data of the target device, thereby ensuring data security.

[0052] Specifically, the non-shared data is substituted into the feature data probability distribution model, and the value output by the feature data probability distribution model is the target device distribution probability of the non-shared data.

[0053] Specifically, the non-shared data is substituted into the probability distribution model of the detection point data, and the value output by the probability distribution model of the detection point data is the detection point distribution probability of the shared data.

[0054] Step 104: Establish a federated learning model according to the non-shared data, the weight of the non-shared data, and the device fault label corresponding to the non-shared data.

[0055] In this embodiment, a federated learning model is established through non-shared data, the weight of non-shared data and the device fault label corresponding to the non-shared data, so as to realize the migration of non-shared data to the target device without sharing the characteristic data of the target device, thereby ensuring data security.

[0056] It should be noted that the characteristic data of the target device and the detection point data are distributed at different detection points in the Internet of Things. Sharing data for model training will cause data security issues. The embodiment of the present invention uses the non-shared data in the detection point data, the weight of the non-shared data, and the device fault label corresponding to the non-shared data to migrate the non-shared data to the target device, so that there is no data sharing between the detection points, avoiding the data security issues caused by directly sharing data. Among them, the detection point is a node that can perform data processing and data interaction, including but not limited to any one or more of an edge server, an edge gateway, and an edge controller.

[0057] Specifically, the equipment fault label can be the characteristic information of the equipment operating status to predict the probability of equipment failure. For example, the equipment fault label can be the fault type or the degree of fault. The embodiment of the present invention does not limit this, and it needs to be determined in combination with actual needs. It should be noted that the embodiment of the present invention does not intend to limit the method for obtaining the equipment fault label corresponding to each non-shared data. It can also be manual labeling, rule labeling, or cluster labeling. The above labeling methods are all prior arts and will not be elaborated here. It is worth noting that the equipment failure prediction method provided in the embodiment of the present invention can also be used to predict the remaining service life of the equipment. Specifically, the equipment fault label can be replaced by the remaining service life of the equipment.

[0058] In one embodiment, establishing a federated learning model according to the non-shared data, the weight of the non-shared data, and the device fault label corresponding to the non-shared data includes:

[0059] Determine a detection point local model according to the non-shared data, the weight of the non-shared data, and the device fault label corresponding to the non-shared data;

[0060] A federated learning model is established based on at least two of the detection point local models.

[0061] Specifically, model training is performed based on non-shared data, weights corresponding to the non-shared data, and equipment fault labels to determine the local model of the detection point. Then, multiple local models of the detection points are fused to establish a federated learning model.

[0062] Specifically, non-shared data and the weight of non-shared data can realize the migration of non-shared data of the detection point to the target device. Subsequently, the model parameters are adjusted by the weights of multiple non-shared data, so that the adjusted model parameters can reflect the relationship between the characteristic data of the target device and the device failure, without involving the sharing of the detection point data and the characteristic data of the target device, thereby ensuring data security.

[0063] In one embodiment of the present invention, establishing a federated learning model based on at least two of the detection point local models includes:

[0064] Based on the joint learning algorithm, at least two of the detection point local models are repeatedly iterated, and a federated learning model is established based on the iterated detection point local models of at least two of the detection points.

[0065] Specifically, the federated learning algorithm refers to sending the local model parameters of each detection point in the local model of each detection point to the detection point of the target device. The detection point of the target device averages or weighted averages the local model parameters of each detection point to obtain the federated learning model parameters, and updates the local model of the detection point based on the federated learning parameters.

[0066] Specifically, by sending the model parameters of the local models of the detection points of each detection point to the detection point of the target device, the detection point of the target device fuses the model parameters of the local models of the detection points from each detection point to obtain the federated learning model parameters, and redistributes the fused federated learning model parameters to the local models of the detection points of each detection point, and then performs local training based on the non-shared data of each detection point and the fused joint model parameters to obtain updated model parameters, and sends the updated model parameters to the detection point of the target device, and iterates repeatedly according to the above method until the number of iterations meets the preset number of times, or the model error of the local model of the iterated detection point meets the preset value. The iterated local model parameters of the detection points are fused to obtain the federated learning model parameters, and then the federated learning model is obtained. Among them, the fusion method of the model parameters of the local models of each detection point can be average or weighted average, which is not limited in the embodiment of the present invention.

[0067] Specifically, the local model of the detection point can be a neural network model or a regression model, which needs to be determined in combination with actual needs.

[0068] Specifically, the local model of the detection point can be iterated in the following way:

[0069] A1. Perform model training based on multiple non-shared data, device fault labels corresponding to each non-shared data, and weights corresponding to each non-shared data to determine a local model of the detection point;

[0070] A2, judging whether the model error of the local model of the detection point meets the iteration condition, if so, determining the local model of the detection point as the final model and sending it to the detection point of the target device, if not, executing A3;

[0071] A3, sending the model parameters of the local model of the detection point to the detection point of the target device;

[0072] A4. Receive the fused model parameters sent by the detection point of the target device, and adjust the fused model parameters according to the multiple non-shared data, the device fault labels corresponding to the multiple non-shared data, and the weights corresponding to the multiple non-shared data to determine the adjusted model parameters, and replace the model parameters of the local model of the detection point with the adjusted model parameters, and execute A2.

[0073] Step S105: Establish a federated learning model based on the local models of at least two detection points.

[0074] Specifically, the current feature data of the target device is collected and substituted into the federated learning model to determine whether the target device is faulty.

[0075] In order to better understand the data processing process between the detection point of the target device and the detection point of the non-shared data, for example, assume that the three detection points of non-shared data are set to A, B and C, and the node of the target device is set to D. Taking A as an example, A determines the probability distribution model of the non-shared data based on the mixed Gaussian model, and receives the probability distribution model of the feature data sent by D, so as to determine the weight corresponding to the non-shared data. B and C follow the similar processing process of A to obtain the weight corresponding to each non-shared data in B and C. After that, A, B, and C perform model training according to each non-shared data, the device fault label corresponding to the non-shared data, and the weight corresponding to each non-shared data, respectively, to obtain the local model of the detection point. After that, D performs federated learning on the local models of the detection points sent by A, B, and C to obtain the federated learning model of the target device.

[0076] It can be seen from the above technical solutions that the embodiments of the present invention have at least the following effective effects: the weight of non-shared data is determined by the probability distribution model of feature data and the probability distribution model of non-shared data to obtain a local model of the detection point of the target device, a data association between the detection point of the non-shared data and the target device is established, and the non-shared data is migrated to the target device. At the same time, a federated learning model is established based on the non-shared data, and the device failure prediction of the target device is realized by using the federated learning model. There is no need to share the feature data of the device between the non-target device and the target device, thereby ensuring data security.

[0077] Figure 1 What is shown is only a basic embodiment of the method of the present invention. By performing certain optimization and expansion on this basis, other preferred embodiments of the method can be obtained.

[0078] In order to more clearly illustrate the technical solution of the present invention, please refer to Figure 2 The embodiment of the present invention provides another device failure prediction method. This embodiment is based on the above embodiment and is further described in combination with a specific application scenario. In this embodiment, the following steps may be specifically included:

[0079] Step 201: Determine feature information of a target device and detection point data information corresponding to the target device.

[0080] Specifically, assume that the target device corresponds to a detection point n in the Internet of Things, and there are N detection points with non-shared data, which correspond to detection points n1, n2, ..., nN in the Internet of Things respectively. n1, n2, ..., nN can obtain non-shared data respectively, and n can obtain all feature data of the target device. It should be noted that the process of obtaining the local model of the detection point n1, n2, ..., nN is similar, and the following only describes the processing process of n1.

[0081] Step 202: Based on the feature information of the target device and the mixed Gaussian model, calculate the data distribution of the feature data corresponding to the feature information, and determine the mixed Gaussian model with determined model parameters as the feature data probability distribution model.

[0082] nAccording to the Gaussian mixture model, each feature data is substituted into the mixed Gaussian model to calculate each feature data, so as to obtain a probability distribution model of each feature data, and each feature data probability distribution model is sent to n1, n2, ..., nN respectively.

[0083] The process of obtaining the probability distribution model of non-shared data for n1, n2, ..., nN is similar. The following describes the process of n1.

[0084] Step 203: Calculate the data distribution of the detection point data with non-shared data corresponding to the detection point data information based on the detection point data information corresponding to the target device and the mixed Gaussian model, and determine the mixed Gaussian model with determined model parameters as the probability distribution model of the non-shared data.

[0085] n1 obtains multiple non-shared data, substitutes the non-shared data into the mixed Gaussian model for calculation, and obtains the probability distribution model of the non-shared data of n1.

[0086] Step 204: Determine the target device distribution probability of the non-shared data according to the probability distribution model of the feature data, and determine the detection point distribution probability of the non-shared data according to the probability distribution model of the detection point data.

[0087] n1 receives the probability distribution model corresponding to the feature data, substitutes the non-shared data it obtains into the probability distribution model of the feature data for calculation, and obtains the target device distribution probability of the non-shared data.

[0088] n1 substitutes the non-shared data into the probability distribution model of the non-shared data for calculation, and obtains the detection point distribution probability of the non-shared data.

[0089] Step 205: Determine the ratio of the target device distribution probability of the non-shared data to the detection point distribution probability of the non-shared data as the weight of the non-shared data.

[0090] n1 determines the weight of non-shared data as the probability ratio of the distribution probability of non-shared data relative to the target device and the distribution probability of the detection point relative to the non-shared data. Based on the same method, the weight of each non-shared data is obtained, so that the data connection between the target device and the detection point of the non-shared data can be established.

[0091] Step 206: Determine the detection point local model according to the non-shared data, the weight of the non-shared data, and the device fault label corresponding to the non-shared data.

[0092] n1 performs model training based on non-shared data, the weight of non-shared data, and the device fault label corresponding to the non-shared data to determine the local model of the detection point of the target device. n2, ..., nN respectively determine the local model of the detection point of the target device using a method similar to n1.

[0093] Step 207: Based on the joint learning algorithm, repeatedly iterate the detection point local models of at least two of the detection points, and establish a federated learning model based on the iterated detection point local models of at least two of the detection points.

[0094] n and n1, n2, ..., nN, use a joint learning algorithm to iterate the local models of the detection points trained by n1, n2, ..., nN respectively, and then send them to n. n averages the iterated local models of the detection points trained by n1, n2, ..., nN respectively, and establishes a federated learning model of the target device.

[0095] Step 208: Use the federated learning model to predict device failures of the target device.

[0096] nPredict equipment failures of target devices based on the federated learning model.

[0097] It can be seen from the above technical scheme that the beneficial effects of this embodiment are: the weight of the non-shared data is obtained by the ratio of the probability distribution model of the non-shared data to the probability distribution model of the feature data, and the data association between the detection point of the non-shared data and the target device is established, and the non-shared data is migrated to the target device. The establishment of a federated learning model of the target device is realized through the target device and the detection point of the non-shared data in the Internet of Things, thereby realizing that the target device and the detection point of the non-shared data do not involve the sharing of feature data, thereby ensuring data security.

[0098] Based on the same concept as the device failure prediction method provided by the method embodiment of the present invention, please refer to Figure 3 The present invention implements a device for predicting equipment status failures, comprising:

[0099] An information determination module 301 is used to determine characteristic information of a target device and detection point data information corresponding to the target device;

[0100] A probability model determination module 302, configured to determine a probability distribution model of characteristic data of the target device and a probability distribution model of detection point data with non-shared data based on the characteristic information of the target device and the detection point data information corresponding to the target device;

[0101] A weight determination module 303, configured to determine the weight of the non-shared data according to the probability distribution model of the feature data and the probability distribution model of the detection point data;

[0102] A model building module 304, configured to build a federated learning model according to the non-shared data, the weight of the non-shared data, and the device fault label corresponding to the non-shared data;

[0103] The prediction module 305 is used to predict the equipment failure of the target equipment according to the federated learning model.

[0104] In one embodiment of the present invention, the probability model determination module 302 includes: a data determination unit, a first probability distribution model determination unit, and a second probability distribution model determination unit; wherein,

[0105] a data determination unit, configured to determine the characteristic data of the target device and the detection point data having non-shared data based on the characteristic information of the target device and the detection point data information corresponding to the target device;

[0106] A first probability distribution model determining unit, configured to determine feature data of the target device and detection point data with non-shared data based on feature information of the target device and detection point data information corresponding to the target device;

[0107] The second probability distribution model determination unit is used to calculate the data distribution of the detection point data according to the detection point data parameter model, and determine the detection point data parameter model with determined parameters as the probability distribution model of the detection point data.

[0108] In this embodiment, the feature data parameter model and / or the detection point data parameter model includes a mixed Gaussian model.

[0109] In one embodiment of the present invention, the weight determination module 303 includes: a first distribution probability determination unit, a second distribution probability unit and a weight determination unit; wherein,

[0110] The first distribution probability determination unit is used to determine the target device distribution probability of the non-shared data according to the characteristic data probability distribution model;

[0111] The second distribution probability unit is used to determine the detection point distribution probability of the non-shared data according to the probability distribution model of the detection point data;

[0112] The weight determination unit is used to determine the ratio of the target device distribution probability of the non-shared data to the detection point distribution probability of the non-shared data as the weight of the non-shared data.

[0113] In one embodiment of the present invention, the model building module 304 includes: a local model determination unit and a model building unit;

[0114] The local model determination unit is used to determine the detection point local model according to the non-shared data, the weight of the non-shared data and the device fault label corresponding to the non-shared data;

[0115] The model building unit is used to build a federated learning model according to the detection point local models of at least two of the detection points.

[0116] In one embodiment of the present invention, the model building unit is used to repeatedly iterate the detection point local models of at least two of the detection points based on a joint learning algorithm, and to build a federated learning model based on the iterated detection point local models of at least two of the detection points.

[0117] In one embodiment of the present invention, the detection point local model includes a neural network model or a regression model.

[0118] Figure 4 It 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 401 and a memory 402 storing execution instructions, and optionally also includes an internal bus 403 and a network interface 404. Among them, the memory 402 may include a memory 4021, such as a high-speed random access memory (Random-Access Memory, RAM), and may also include a non-volatile memory 4022 (non-volatile memory), such as at least one disk storage, etc.; the processor 401, the network interface 404 and the memory 402 can be interconnected through an internal bus 403, and the internal bus 403 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 403 can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 4 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 501 executes the execution instruction stored in the memory 502, the processor 401 executes the method in any embodiment of the present invention, and is at least used to execute the following steps: Figure 1 or Figure 2 The method shown.

[0119] In a possible implementation, the processor reads the corresponding execution instruction from the non-volatile memory into the memory and then runs it, and can also obtain the corresponding execution instruction from other devices to form a device fault prediction device at the logical level. The processor executes the execution instruction stored in the memory to implement a device fault prediction method provided in any embodiment of the present invention through the execution of the execution instruction.

[0120] 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.

[0121] 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 4 The electronic device shown; the execution instruction is a computer program corresponding to a device failure prediction method.

[0122] 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.

[0123] The multiple embodiments of the present invention are described in a progressive manner, and the same or similar parts between the multiple 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.

[0124] 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 device 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 device. 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 device including the elements.

[0125] 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 predicting equipment failure, characterized in that: include: Determine characteristic information of the target device and detection point data information corresponding to the target device; The detection point is a node in the Internet of Things that can perform data processing and data interaction; Determine a probability distribution model of the characteristic data of the target device and a probability distribution model of the detection point data with non-shared data based on the characteristic information of the target device and the detection point data information corresponding to the target device; Determining the weight of the non-shared data according to the probability distribution model of the feature data and the probability distribution model of the detection point data; Establishing a federated learning model according to the non-shared data, the weight of the non-shared data, and the device fault label corresponding to the non-shared data; Performing equipment failure prediction of the target equipment according to the federated learning model; The determining the weight of the non-shared data according to the probability distribution model of the feature data and the probability distribution model of the detection point data includes: Determining a target device distribution probability of the non-shared data according to the characteristic data probability distribution model; Determining the detection point distribution probability of the non-shared data according to the probability distribution model of the detection point data; Determine a ratio of a target device distribution probability of the non-shared data to a detection point distribution probability of the non-shared data as a weight of the non-shared data; The establishing of a federated learning model according to the non-shared data, the weight of the non-shared data, and the device fault label corresponding to the non-shared data includes: Determine a detection point local model according to the non-shared data, the weight of the non-shared data, and the device fault label corresponding to the non-shared data; A federated learning model is established based on at least two of the detection point local models.

2. The method according to claim 1, characterized in that The determining, based on the feature information of the target device and the detection point data information corresponding to the target device, a probability distribution model of the feature data of the target device and a probability distribution model of the detection point data with non-shared data comprises: Determining the characteristic data of the target device and the detection point data with non-shared data based on the characteristic information of the target device and the detection point data information corresponding to the target device; Calculate the data distribution of the characteristic data according to the characteristic data parameter model, and determine the characteristic data parameter model with determined parameters as the characteristic data probability distribution model; The data distribution of the detection point data is calculated according to the detection point data parameter model, and the detection point data parameter model with determined parameters is determined as the probability distribution model of the detection point data.

3. The method according to claim 2, characterized in that The feature data parameter model and / or the detection point data parameter model comprises a mixed Gaussian model.

4. The method according to claim 1, characterized in that: The step of establishing a federated learning model based on at least two of the detection point local models includes: Based on the joint learning algorithm, at least two of the detection point local models are repeatedly iterated, and a federated learning model is established based on at least two iterated detection point local models.

5. The method according to claim 1, characterized in that The detection point local model includes a neural network model or a regression model.

6. A device for predicting equipment failure, characterized in that: include: An information determination module, used to determine characteristic information of a target device and detection point data information corresponding to the target device; The detection point is a node in the Internet of Things that can perform data processing and data interaction; A probability model determination module, used to determine a probability distribution model of characteristic data of the target device and a probability distribution model of detection point data with non-shared data based on the characteristic information of the target device and the detection point data information corresponding to the target device; A weight determination module, used to determine the weight of the non-shared data according to the probability distribution model of the feature data and the probability distribution model of the detection point data; A model building module, used to build a federated learning model according to the non-shared data, the weight of the non-shared data and the device fault label corresponding to the non-shared data; A prediction module, used for predicting equipment failure of the target equipment according to the federated learning model; The weight determination module is specifically used to: determine the target device distribution probability of the non-shared data according to the feature data probability distribution model; determine the detection point distribution probability of the non-shared data according to the detection point data probability distribution model; determine the ratio of the target device distribution probability of the non-shared data to the detection point distribution probability of the non-shared data as the weight of the non-shared data; The model building module is specifically used to: determine the detection point local model according to the non-shared data, the weight of the non-shared data and the equipment fault label corresponding to the non-shared data; and establish a federated learning model according to at least two of the detection point local models.

7. 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 5.

8. 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 5.

Images