Method and apparatus for predicting appliance failure based on neuron model

Abstract

Embodiments of the present application provide a kind of electric appliance fault prediction method and device based on neuron model, electronic equipment and readable storage medium, the method comprises: constructing neuron model and weight regression model, determine the business weight mark and nerve node mark of the neuron model;Obtain the fault diagnosis sample data of specified electric appliance, and according to the fault diagnosis sample data of the specified electric appliance, the business weight mark and the nerve node mark, the neuron model is trained;Using the weight regression model determines the fitting value of the neuron model after training;When the fitting value meets preset threshold value, the neuron model after training is determined as the neuron model of training completion, to carry out fault prediction according to the specified electric appliance of the neuron model of training completion.This embodiment of the present application can avoid the information collected by the electric appliance of fault to cause intelligent perception error, improve the accuracy of intelligent perception.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of intelligent sensing technology, and in particular to an electrical fault prediction method based on a neuron model, an electrical fault prediction device based on a neuron model, an electronic device, and a computer-readable storage medium. Background Technology

[0002] Information technology is a crucial indicator of a nation's comprehensive national strength and one of the most fiercely competitive fields in the world today. Among these, intelligent sensing is a key cutting-edge technology that is receiving significant development in the field of information technology.

[0003] With the rapid development of the Internet of Things (IoT) and its widespread application across various industries, a large amount of high-dimensional information has emerged in intelligent sensing. This high-dimensional information is characterized by its high feature dimensionality, massive feature quantity, and the presence of a large amount of irrelevant and redundant information. Continuous exploration and research are needed to address the challenges of intelligent sensing with this high-dimensional information. Specifically, key technologies for high-dimensional information research include electrical appliances such as sensors and transformers. To achieve accurate intelligent sensing, it is necessary to predict the electrical faults of IoT devices to avoid errors in intelligent sensing caused by using information collected from faulty devices. Summary of the Invention

[0004] The present invention provides an electrical appliance fault prediction method, device, electronic device, and computer-readable storage medium based on a neuron model, in order to avoid the problem of intelligent sensing errors caused by using information collected from faulty electrical appliances.

[0005] This invention discloses an electrical fault prediction method based on a neuron model, comprising:

[0006] Construct a neuron model and a weighted regression model, wherein the neuron model includes at least an input layer, a hidden layer, and an output layer;

[0007] Determine the service weight identifier and the neural node identifier of the neuron model; wherein, the service weight identifier and the neural node identifier are the data brought in when the data of the input layer is processed into the hidden layer;

[0008] Obtain fault diagnosis sample data of a specified electrical appliance, and train the neuron model based on the fault diagnosis sample data of the specified electrical appliance, the business weight identifier, and the neural node identifier;

[0009] The weighted regression model is used to determine the fitted values ​​of the input layer and the output layer of the trained neuron model;

[0010] When the fitted value meets the preset threshold, the trained neuron model is determined as the trained neuron model, so as to predict the fault of the specified electrical appliance based on the trained neuron model.

[0011] Optionally, after determining the fitted values ​​of the input layer and the output layer of the trained neuron model using the weighted regression model, the method further includes:

[0012] When the fitted value does not meet the preset threshold, the business weight identifier and the neural node identifier are adjusted, and the process returns to the step of using the weight regression model to determine the fitted values ​​of the input layer and the output layer of the trained neuron model.

[0013] Optionally, determining the business weight identifier and neural node identifier of the neuron model includes:

[0014] The business weight identifier is determined based on the historical database topology of the neuron model, and the neural node identifier is determined based on the historical data of the same type in the neuron model.

[0015] When it is impossible to determine the business weight identifier based on the historical database topology of the neuron model, and it is impossible to determine the neural node identifier based on the historical data of the same type of neuron model, the business weight identifier and the neural node identifier are randomly assigned.

[0016] Optionally, the hidden layer includes multiple neurons. After acquiring fault diagnosis sample data of a specified appliance and training the neuron model based on the fault diagnosis sample data of the specified appliance, the business weight identifier, and the neural node identifier, the method further includes:

[0017] Determine the difference between the number of neurons currently allocated in the hidden layer and the number of neurons allocated in the previous time;

[0018] When the differentiated data exceeds a preset differentiated threshold, the business weight identifier and the neural node identifier are adjusted.

[0019] Optionally, the weighted regression model includes a ridge regression model, the formula of which is:

[0020] ||Xθ-y|| 2 +||Γθ|| 2

[0021] Wherein, X represents the input data, y represents the output data, Γ represents the training result of the neuron model, and θ is the fitting hyperparameter;

[0022] The given Γ=aI, θ(a=(X) T X+aI) -1 X T y

[0023] Wherein, I represents the identity matrix, T is a weight constant, a is the weight of the identity matrix I, and θ(a) represents the calculation of θ given a.

[0024] This invention also discloses an electrical fault prediction device based on a neuron model, comprising:

[0025] The model building module is used to build a neuron model and a weighted regression model, wherein the neuron model includes at least an input layer, a hidden layer and an output layer;

[0026] The identifier determination module is used to determine the business weight identifier and the neural node identifier of the neuron model; wherein, the business weight identifier and the neural node identifier are the data brought in when the data of the input layer is processed into the hidden layer;

[0027] The model training module is used to acquire fault diagnosis sample data of a specified electrical appliance, and train the neuron model based on the fault diagnosis sample data of the specified electrical appliance, the business weight identifier, and the neural node identifier.

[0028] The fitting value acquisition module is used to determine the fitting values ​​of the input layer and the output layer of the trained neuron model using the weighted regression model;

[0029] The training determination module is used to determine the trained neuron model as the trained neuron model when the fitted value meets a preset threshold, so as to predict the fault of the specified electrical appliance based on the trained neuron model.

[0030] Optionally, the apparatus further includes a return execution module for:

[0031] When the fitted value does not meet the preset threshold, the business weight identifier and the neural node identifier are adjusted, and the process returns to the step of using the weight regression model to determine the fitted values ​​of the input layer and the output layer of the trained neuron model.

[0032] Optionally, the identifier determination module is specifically used for:

[0033] The business weight identifier is determined based on the historical database topology of the neuron model, and the neural node identifier is determined based on the historical data of the same type in the neuron model.

[0034] When it is impossible to determine the business weight identifier based on the historical database topology of the neuron model, and it is impossible to determine the neural node identifier based on the historical data of the same type of neuron model, the business weight identifier and the neural node identifier are randomly assigned.

[0035] Optionally, the hidden layer includes multiple neurons, and the device further includes an identifier adjustment module for:

[0036] Determine the difference between the number of neurons currently allocated in the hidden layer and the number of neurons allocated in the previous time;

[0037] When the differentiated data exceeds a preset differentiated threshold, the business weight identifier and the neural node identifier are adjusted.

[0038] Optionally, the weighted regression model includes a ridge regression model, the formula of which is:

[0039] ||Xθ-y|| 2 +||Γθ|| 2

[0040] Wherein, X represents the input data, y represents the output data, Γ represents the training result of the neuron model, and θ is the fitting hyperparameter;

[0041] The given Γ=aI, θ(a=(X) T X+aI) -1 X T y

[0042] Wherein, I represents the identity matrix, T is a weight constant, a is the weight of the identity matrix I, and θ(a) represents the calculation of θ given a.

[0043] This invention also discloses an electronic device, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus;

[0044] The memory is used to store computer programs;

[0045] When the processor executes a program stored in the memory, it implements the method described in the embodiments of the present invention.

[0046] This invention also discloses a computer program product stored in a storage medium, which is executed by at least one processor to implement the method described in this invention.

[0047] This invention also discloses a computer-readable storage medium storing instructions that, when executed by one or more processors, cause the processors to perform the methods described in this invention.

[0048] The embodiments of the present invention have the following advantages:

[0049] In this embodiment of the invention, a neuron model and a weighted regression model are constructed. The neuron model may include at least an input layer, a hidden layer, and an output layer. The business weight identifier and neural node identifier of the neuron model are determined. The business weight identifier and neural node identifier are data introduced into the hidden layer when the data from the input layer is processed. Then, fault diagnosis sample data of a specified appliance can be obtained. The neuron model is trained based on the fault diagnosis sample data of the specified appliance, the business weight identifier, and the neural node identifier. The weighted regression model is then used to determine the fitting values ​​of the input and output layers of the trained neuron model. When the fitting value meets a preset threshold, the trained neuron model can be identified as a completed neuron model. This completed neuron model can predict faults in specified appliances in the Internet of Things, thereby avoiding errors in intelligent sensing caused by using information collected from faulty appliances and improving the accuracy of intelligent sensing. Attached Figure Description

[0050] Figure 1 This is a flowchart illustrating the steps of an electrical fault prediction method based on a neuron model provided in an embodiment of the present invention.

[0051] Figure 2 This is a schematic diagram of a neuron model provided in an embodiment of the present invention;

[0052] Figure 3 This is a structural block diagram of an electrical fault prediction device based on a neuron model provided in an embodiment of the present invention;

[0053] Figure 4 This is a schematic diagram of the hardware structure of an electronic device that implements various embodiments of the present invention. Detailed Implementation

[0054] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0055] Reference Figure 1 The diagram illustrates a flowchart of an electrical fault prediction method based on a neuron model provided in an embodiment of the present invention, which may specifically include the following steps:

[0056] Step 101: Construct a neuron model and a weighted regression model, wherein the neuron model includes at least an input layer, a hidden layer, and an output layer.

[0057] The embodiments of the present invention are applicable to the field of intelligent sensing in electrical engineering. The electrical appliances can include various electrical appliances in the Internet of Things, such as sensors, transformers, etc. The embodiments of the present invention do not impose any limitations on this.

[0058] In practical implementation, the neuron model is constructed from a neural network. Specifically, a neural network is composed of many neurons, and the neuron is the most basic unit of the neural network. The neural network structure for constructing the neuron model in this embodiment of the invention can be categorized at least as follows: feedforward neural network, output feedback feedforward network, feedforward inner interconnection network, feedback fully interconnection network, and feedback locally interconnection network.

[0059] Specifically, in feedforward neural networks, the input is transmitted layer by layer along the network, with no feedback between layers. Common examples include neural networks and orthogonal networks. Feedforward networks with output feedback have feedback from the output, meaning that the input signal of a neuron can come from both external input and feedback from its own output. Feedforward inner-layer interconnected networks are characterized by interconnected and mutually restrictive neurons within each layer, with no feedback between layers. This type of neural network is often a self-organizing neural network. Feedback-type fully interconnected networks are more complex, characterized by the fact that the output of any neuron is connected to other neurons. Such networks are feedback-type fully interconnected networks. Feedback-type locally interconnected networks do not have connections between each neuron and all surrounding neurons. They are commonly used in the identification of nonlinear systems.

[0060] In this embodiment of the invention, in order to accurately predict (diagnose) faults in a specified electrical appliance, such as a sensor or transformer, a neuron model and a weighted regression model can be constructed first. For example, refer to... Figure 2 This is a schematic diagram of a neuron model provided in an embodiment of the present invention. The neuron model includes at least an input layer, a hidden layer, and an output layer, wherein the hidden layer may include multiple neurons.

[0061] As an optional example, the neuron model can be:

[0062] The transformation function f is:

[0063] Local nonlinearity can be divided into threshold-type and piecewise linear type. The threshold-type is as follows:

[0064] f(x i )=1,x i >0OR f(x) i )=1,x i ≤0

[0065] The piecewise linear form is:

[0066] f(x i ) = 0, x i ≤x i0 OR f(x i )=kx i x i0 <x i <x i1

[0067] OR f(x i )=f max x i ≥x i

[0068] Commonly used conversion functions include the Sigmoid function and the Tan function. The Sigmoid function is as follows:

[0069]

[0070] The Tan function is:

[0071]

[0072] Of course, the above-described neuron model is merely an example. Other neuron models can be used according to actual needs when implementing the embodiments of the present invention, and the embodiments of the present invention do not impose any restrictions on this.

[0073] Step 102: Determine the business weight identifier and neural node identifier of the neuron model; wherein, the business weight identifier and the neural node identifier are the data brought in when the data of the input layer is processed into the hidden layer.

[0074] In practical implementation, the number of neurons in the hidden layers of a backpropagation (BP) neural network directly affects the convergence speed and diagnostic accuracy. Insufficient neurons result in insufficient information acquisition, potentially causing the network to get stuck in local minima or fail to achieve training results. Conversely, excessive neurons can lead to overfitting and prolonged training time. Therefore, determining the number of neurons in the neural network model is crucial. This number can be determined by the business weight identifier and the neural node identifier. Choosing appropriate business weight identifiers and neural node identifiers is essential for training the neural network model, thereby improving training effectiveness and achieving better training results.

[0075] In an optional embodiment of the present invention, step 102, determining the service weight identifier and neural node identifier of the neuron model, may include:

[0076] The business weight identifier is determined based on the historical database topology of the neuron model, and the neural node identifier is determined based on the historical data of the same type in the neuron model.

[0077] When it is impossible to determine the business weight identifier based on the historical database topology of the neuron model, and it is impossible to determine the neural node identifier based on the historical data of the same type of neuron model, the business weight identifier and the neural node identifier are randomly assigned.

[0078] In this embodiment of the invention, before the input data enters the hidden layer of the neuron model, the business weight identifier can be determined by combining the historical database topology relationship. Then, the data after entering the hidden layer of the neuron model is equipped with the business weight identifier for training the neuron model. Of course, if the business weight identifier cannot be determined according to the historical data topology relationship of the neuron model, an initial business weight identifier can be randomly assigned.

[0079] Meanwhile, in this embodiment of the invention, before the input data enters the hidden layer of the neuron model, the neural node identifiers assigned in the hidden layer operations of a batch of historical data of the same type are obtained. Then, the neural node identifiers are used to train the neuron model after the data enters the hidden layer. Of course, if it is not possible to obtain the neural node identifiers of historical data of the same type, an initial neural node identifier can be randomly assigned.

[0080] Therefore, if historical business weight identifiers and neural node identifiers can be obtained before training the neuron model, they can be used for training. Since these historical business weight identifiers and neural node identifiers are related to the neuron model, using them can improve the training efficiency. Of course, even if historical business weight identifiers and neural node identifiers cannot be obtained, they can be randomly assigned to the neuron model for training.

[0081] Step 103: Obtain fault diagnosis sample data of the specified electrical appliance, and train the neuron model based on the fault diagnosis sample data of the specified electrical appliance, the business weight identifier, and the neural node identifier.

[0082] Step 104: Use the weighted regression model to determine the fitting values ​​of the input layer and the output layer of the trained neuron model.

[0083] As an optional example, the weighted regression model may include a ridge regression method model, the formula of which can be:

[0084] ||Xθ-y|| 2 +||Γθ|| 2

[0085] Wherein, X represents the input data, y represents the output data, Γ represents the training result of the neuron model, and θ is the fitting hyperparameter;

[0086] The given Γ=aI, θ(a=(X) T X+aI) -1 X T y

[0087] Wherein, I represents the identity matrix, T is a weight constant, a is the weight of the identity matrix I, and θ(a) represents the calculation of θ given a.

[0088] Of course, in addition to the ridge regression method model mentioned above as a weighted regression model, other algorithms or models can also be selected in the embodiments of the present invention, and the embodiments of the present invention do not impose any restrictions on this.

[0089] In this embodiment of the invention, after constructing the neuron model and obtaining the business weight identifier and the neural node identifier, fault diagnosis sample data of the specified electrical appliance can be obtained. Then, the neuron model can be trained based on the fault diagnosis sample data of the specified electrical appliance, the business weight identifier, and the neural node identifier. Furthermore, in order to determine the accuracy of the trained neuron model, a weighted regression model can be further used to determine the fitting values ​​of the input layer and the output layer of the trained neuron model, so as to determine whether the training of the neuron model is completed based on the fitting values.

[0090] Step 105: When the fitted value meets the preset threshold, the trained neuron model is determined as the trained neuron model, so as to predict the fault of the specified electrical appliance based on the trained neuron model.

[0091] In this embodiment of the invention, when the fitting values ​​of the input layer and the output layer of the neuron model meet the preset threshold, the trained neuron model can be determined as the trained neuron model, and the neuron model can be used to predict the faults of specified electrical appliances in the Internet of Things.

[0092] In the aforementioned electrical appliance fault prediction method based on a neuron model, a neuron model and a weighted regression model are constructed. The neuron model can include at least an input layer, a hidden layer, and an output layer. The business weight identifier and neural node identifier of the neuron model are determined. These identifiers are the data brought into the hidden layer when the input layer data is processed. Then, fault diagnosis sample data of a specified electrical appliance can be obtained. The neuron model is trained based on the fault diagnosis sample data, business weight identifier, and neural node identifier. The weighted regression model is then used to determine the fitting values ​​of the input and output layers of the trained neuron model. When the fitting values ​​meet a preset threshold, the trained neuron model is considered a completed neuron model. This completed neuron model can predict faults in specified electrical appliances in the Internet of Things (IoT), thereby avoiding intelligent sensing errors caused by using information collected from faulty appliances and improving the accuracy of intelligent sensing.

[0093] In a preferred embodiment of the present invention, after step 104, which involves determining the fitting values ​​of the input layer and the output layer of the trained neuron model using the weighted regression model, the method may further include:

[0094] When the fitted value does not meet the preset threshold, the business weight identifier and the neural node identifier are adjusted, and the process returns to the step of using the weight regression model to determine the fitted values ​​of the input layer and the output layer of the trained neuron model.

[0095] In this embodiment of the invention, if the fitting values ​​of the input layer and output layer of the neuron model do not meet the preset threshold, the business weight identifier and the neural node identifier can be adjusted, and the neuron model can be retrained using the adjusted business weight identifier and neural node identifier until the fitting value of the neuron model meets the preset threshold, which can ensure the accuracy of the neuron model in predicting electrical faults.

[0096] In a preferred embodiment of the present invention, after step 103, obtaining fault diagnosis sample data of a specified electrical appliance, and training the neuron model based on the fault diagnosis sample data of the specified electrical appliance, the business weight identifier, and the neural node identifier, the method may further include:

[0097] Determine the difference between the number of neurons currently allocated in the hidden layer and the number of neurons allocated in the previous time;

[0098] When the differentiated data exceeds a preset differentiated threshold, the business weight identifier and the neural node identifier are adjusted.

[0099] In this process, the computation results of the hidden layer of the neuron model are passed to the output layer. The output layer obtains the number of neurons (usually called neural nodes) allocated in the current hidden layer computation, compares it with the number of neurons allocated in the previous hidden layer computation, and then determines the differential data between the two. When the differential data exceeds the preset differential threshold (e.g., a difference of 10%), two thresholds need to be adjusted, namely the business weight identifier and the neural node identifier. Then, the data is put into the hidden layer for training in the next training of the neuron model. This reduces the initial computational consumption of the hidden layer resources after the data is put into the hidden layer, and completes the detection of the effective number of fault diagnosis sample data of the specified electrical appliance.

[0100] To enable those skilled in the art to better understand the embodiments of the present invention, a complete specific example is used for illustration below.

[0101] Step 1: Constructing the Neuron Model. A neural network in a neuron model is composed of many neurons, which are the most basic unit of a neural network. When data enters the hidden layer of the neuron model, the hidden layer performs random business weight assignment and neuron node assignment operations (or neuron node threshold assignment) based on an initial weight allocation threshold.

[0102] S1. The hidden layer acts as an accelerator for neural network operations, and its performance is primarily influenced by the number of neurons and the rationality of randomly assigned thresholds (i.e., business weight labels and neural node labels). Once data enters the hidden layer, it performs random allocation of business weight labels and neural node labels.

[0103] S2. Before the input data enters the hidden layer, the business weight identifier is determined by combining the historical database topology relationship. After entering the hidden layer, the data is brought in for neuron model training. If there is no business weight identifier, an initial business weight identifier is randomly assigned.

[0104] S3. Simultaneously, before the input data enters the hidden layer, obtain the neural node identifiers assigned in the hidden layer operation of the previous batch of historical data of the same type and bring them into the hidden layer. After entering the hidden layer, if the data has a neural node identifier, use that neural node identifier to train the neural model; otherwise, assign an initial neural node identifier.

[0105] S4. The results of the hidden layer's calculations are passed to the output layer. The output layer obtains the number of neurons (neuron nodes) allocated in this hidden layer calculation, compares it with the number of neurons (neuron nodes) allocated in the previous hidden layer calculation, and then determines the differential data between the two. When the differential data exceeds the preset differential threshold, the business weight identifier and the neural node identifier need to be adjusted. Then, the data is put into the hidden layer for training in the next training of the neuron model, thereby reducing the initial computational consumption of the hidden layer resources after the data is put into the hidden layer.

[0106] Step 2: Analyze the effectiveness of training hidden layer data using an AI chip combined with a weighted regression model for the input and output layers of the hidden nodes in a BP neural network.

[0107] By constructing a weighted regression model using the ridge regression method, we can prevent the output data of the hidden layer (reservoir) from overfitting to the simulated input data during machine learning. If overfitting occurs, it is necessary to adjust the threshold of the input layer data or change the random threshold between the neuron nodes in the reservoir calculation.

[0108] Optionally, the weighted regression model includes the ridge regression method model, the formula of which is:

[0109] ||Xθ-y|| 2 +||Γθ|| 2

[0110] Where X represents the input data, y represents the output data, Γ represents the training result of the neuron model, and θ is the fitting hyperparameter;

[0111] The given Γ=aI, θ(a=(X) T X+aI) -1 X T y

[0112] Where I represents the identity matrix, T is a weight constant, a is the weight of the identity matrix I, and θ(a) represents the calculation of θ given a. Specifically, the result of the calculation to prevent overfitting is the fitted value obtained by the data model calculation of the output or input layer. Then, based on the fitted value, it can be determined whether the training of the neuron model can be terminated, so that the trained neuron model can be used to predict the faults of designated electrical appliances in the Internet of Things. Optionally, the training of the neuron model can also be terminated when the maximum number of training iterations is reached.

[0113] Of course, in addition to the ridge regression method model mentioned above, other algorithm models can also be used in this embodiment of the invention to determine whether the neuron model meets the conditions, and this embodiment of the invention does not impose any restrictions on this.

[0114] In summary, by applying the embodiments of the present invention, firstly, a neural network model can be creatively constructed to predict the probability of fault diagnosis for IoT-related electrical appliances, thereby completing the learning method of neural network application for fault diagnosis of IoT-related electrical appliances. Simultaneously, the embodiments of the present invention innovatively optimize the hidden layer neurons of the neural network model using a weighted regression model in an AI chip, combining this with the fault diagnosis method of IoT-related electrical appliances, to improve the training effect of the neural network model.

[0115] It should be noted that, for the sake of simplicity, the method embodiments are all described as a series of actions. However, those skilled in the art should understand that the embodiments of the present invention are not limited to the described order of actions, because according to the embodiments of the present invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions involved are not necessarily essential to the embodiments of the present invention.

[0116] Reference Figure 3 The diagram illustrates a structural block diagram of an electrical fault prediction device based on a neuron model provided in an embodiment of the present invention, which may specifically include the following modules:

[0117] The model building module 301 is used to build a neuron model and a weighted regression model, wherein the neuron model includes at least an input layer, a hidden layer and an output layer;

[0118] The identifier determination module 302 is used to determine the business weight identifier and the neural node identifier of the neuron model; wherein, the business weight identifier and the neural node identifier are the data brought in when the data of the input layer is processed into the hidden layer;

[0119] The model training module 303 is used to acquire fault diagnosis sample data of a specified electrical appliance, and train the neuron model based on the fault diagnosis sample data of the specified electrical appliance, the business weight identifier, and the neural node identifier.

[0120] The fitting value acquisition module 304 is used to determine the fitting values ​​of the input layer and the output layer of the trained neuron model using the weighted regression model;

[0121] The training determination module 305 is used to determine the trained neuron model as the trained neuron model when the fitted value meets a preset threshold, so as to predict the fault of the specified electrical appliance based on the trained neuron model.

[0122] In an optional embodiment of the present invention, the apparatus further includes a return execution module for:

[0123] When the fitted value does not meet the preset threshold, the business weight identifier and the neural node identifier are adjusted, and the process returns to the step of using the weight regression model to determine the fitted values ​​of the input layer and the output layer of the trained neuron model.

[0124] In an optional embodiment of the present invention, the identifier determination module 302 is specifically used for:

[0125] The business weight identifier is determined based on the historical database topology of the neuron model, and the neural node identifier is determined based on the historical data of the same type in the neuron model.

[0126] When it is impossible to determine the business weight identifier based on the historical database topology of the neuron model, and it is impossible to determine the neural node identifier based on the historical data of the same type of neuron model, the business weight identifier and the neural node identifier are randomly assigned.

[0127] In an optional embodiment of the present invention, the device further includes an identification adjustment module for:

[0128] Determine the difference between the number of neurons currently allocated in the hidden layer and the number of neurons allocated in the previous time;

[0129] When the differentiated data exceeds a preset differentiated threshold, the business weight identifier and the neural node identifier are adjusted.

[0130] In an optional embodiment of the present invention, the weighted regression model includes a ridge regression method model, the formula of which is:

[0131] ||Xθ-y|| 2 +||Γθ|| 2

[0132] Wherein, X represents the input data, y represents the output data, Γ represents the training result of the neuron model, and θ is the fitting hyperparameter;

[0133] The given Γ=aI, θ(a=(X) T X+aI) -1 X T y

[0134] Wherein, I represents the identity matrix, T is a weight constant, a is the weight of the identity matrix I, and θ(a) represents the calculation of θ given a.

[0135] In this embodiment of the invention, a neuron model and a weighted regression model are constructed. The neuron model may include at least an input layer, a hidden layer, and an output layer. The business weight identifier and neural node identifier of the neuron model are determined. The business weight identifier and neural node identifier are data introduced into the hidden layer when the data from the input layer is processed. Then, fault diagnosis sample data of a specified appliance can be obtained. The neuron model is trained based on the fault diagnosis sample data of the specified appliance, the business weight identifier, and the neural node identifier. The weighted regression model is then used to determine the fitting values ​​of the input and output layers of the trained neuron model. When the fitting value meets a preset threshold, the trained neuron model can be identified as a completed neuron model. This completed neuron model can predict faults in specified appliances in the Internet of Things, thereby avoiding errors in intelligent sensing caused by using information collected from faulty appliances and improving the accuracy of intelligent sensing.

[0136] As the device embodiment is basically similar to the method embodiment, the description is relatively simple, and relevant parts can be found in the description of the method embodiment.

[0137] In addition, this invention also provides an electronic device, including: a processor, a memory, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, it implements the various processes of the above-described embodiment of the electrical fault prediction method based on the neuron model and achieves the same technical effect. To avoid repetition, it will not be described again here.

[0138] This invention also provides a computer-readable storage medium storing a computer program. When executed by a processor, the computer program implements the various processes of the above-described embodiment of the electrical fault prediction method based on a neuron model, achieving the same technical effects. To avoid repetition, it will not be described again here. The computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0139] This invention also provides a computer program product, which is stored in a storage medium and executed by at least one processor to implement the various processes of the above-described embodiment of the electrical fault prediction method based on the neuron model, and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0140] Figure 4 A schematic diagram of the hardware structure of an electronic device for implementing various embodiments of the present invention.

[0141] The electronic device 400 includes, but is not limited to, components such as: a radio frequency unit 401, a network module 402, an audio output unit 403, an input unit 404, a sensor 405, a display unit 406, a user input unit 407, an interface unit 408, a memory 409, a processor 410, and a power supply 411. Those skilled in the art will understand that... Figure 4 The electronic device structures shown are not intended to limit the electronic device. An electronic device may include more or fewer components than shown, or combine certain components, or have different component arrangements. In embodiments of the present invention, the electronic device includes, but is not limited to, mobile phones, tablet computers, laptops, PDAs, in-vehicle terminals, wearable devices, and pedometers.

[0142] It should be understood that, in this embodiment of the invention, the radio frequency unit 401 can be used for receiving and transmitting signals during information transmission or calls. Specifically, it receives downlink data from the base station and processes it with the processor 410; additionally, it transmits uplink data to the base station. Typically, the radio frequency unit 401 includes, but is not limited to, an antenna, at least one amplifier, a transceiver, a coupler, a low-noise amplifier, a duplexer, etc. Furthermore, the radio frequency unit 401 can also communicate with networks and other devices through a wireless communication system.

[0143] The electronic device provides users with wireless broadband internet access through network module 402, such as helping users send and receive emails, browse web pages, and access streaming media.

[0144] The audio output unit 403 can convert audio data received by the radio frequency unit 401 or the network module 402 or stored in the memory 409 into audio signals and output them as sound. Furthermore, the audio output unit 403 can also provide audio output related to specific functions performed by the electronic device 400 (e.g., call signal reception sound, message reception sound, etc.). The audio output unit 403 includes a speaker, a buzzer, and a receiver, etc.

[0145] Input unit 404 is used to receive audio or video signals. Input unit 404 may include a graphics processing unit (GPU) 4041 and a microphone 4042. The GPU 4041 processes image data of still images or videos acquired by an image capture device (such as a camera) in video capture mode or image capture mode. The processed image frames can be displayed on display unit 406. The image frames processed by GPU 4041 can be stored in memory 409 (or other storage medium) or transmitted via radio frequency unit 401 or network module 402. Microphone 4042 can receive sound and process such sound into audio data. The processed audio data can be converted into a format that can be transmitted to a mobile communication base station via radio frequency unit 401 in telephone call mode.

[0146] The electronic device 400 also includes at least one sensor 405, such as a light sensor, a motion sensor, and other sensors. Specifically, the light sensor includes an ambient light sensor and a proximity sensor. The ambient light sensor can adjust the brightness of the display panel 4061 according to the ambient light level, and the proximity sensor can turn off the display panel 4061 and / or backlight when the electronic device 400 is moved to the ear. As a type of motion sensor, an accelerometer sensor can detect the magnitude of acceleration in various directions (generally three axes). When stationary, it can detect the magnitude and direction of gravity and can be used to identify the posture of the electronic device (such as landscape / portrait switching, related games, magnetometer posture calibration), vibration recognition related functions (such as pedometer, tapping), etc. The sensor 405 may also include a fingerprint sensor, pressure sensor, iris sensor, molecular sensor, gyroscope, barometer, hygrometer, thermometer, infrared sensor, etc., which will not be described in detail here.

[0147] The display unit 406 is used to display information input by the user or information provided to the user. The display unit 406 may include a display panel 4061, which may be configured in the form of a liquid crystal display (LCD), an organic light-emitting diode (OLED), or the like.

[0148] User input unit 407 can be used to receive input numerical or character information, and generate key signal inputs related to user settings and function control of electronic devices. Specifically, user input unit 407 includes a touch panel 4071 and other input devices 4072. Touch panel 4071, also known as a touch screen, can collect touch operations performed by the user on or near it (such as operations performed by the user using a finger, stylus, or any suitable object or accessory on or near touch panel 4071). Touch panel 4071 may include two parts: a touch detection device and a touch controller. The touch detection device detects the user's touch position and the signal generated by the touch operation, and transmits the signal to the touch controller; the touch controller receives touch information from the touch detection device, converts it into touch point coordinates, and sends it to the processor 410, which receives and executes commands from the processor 410. In addition, touch panel 4071 can be implemented using various types such as resistive, capacitive, infrared, and surface acoustic wave. Besides touch panel 4071, user input unit 407 may also include other input devices 4072. Specifically, other input devices 4072 may include, but are not limited to, physical keyboards, function keys (such as volume control buttons, power buttons, etc.), trackballs, mice, joysticks, etc., which will not be described in detail here.

[0149] Furthermore, the touch panel 4071 can cover the display panel 4061. When the touch panel 4071 detects a touch operation on or near it, it transmits the information to the processor 410 to determine the type of touch event. Subsequently, the processor 410 provides corresponding visual output on the display panel 4061 based on the type of touch event. Although in Figure 4 In this embodiment, the touch panel 4071 and the display panel 4061 are two independent components to realize the input and output functions of the electronic device. However, in some embodiments, the touch panel 4071 and the display panel 4061 can be integrated to realize the input and output functions of the electronic device. The specific implementation is not limited here.

[0150] Interface unit 408 serves as an interface for connecting external devices to electronic device 400. For example, external devices may include a wired or wireless headphone port, an external power supply (or battery charger) port, a wired or wireless data port, a memory card port, a port for connecting a device with an identification module, an audio input / output (I / O) port, a video I / O port, a headphone port, and so on. Interface unit 408 can be used to receive input from external devices (e.g., data, power, etc.) and transmit the received input to one or more components within electronic device 400, or it can be used to transmit data between electronic device 400 and external devices.

[0151] The memory 409 can be used to store software programs and various data. The memory 409 may primarily include a program storage area and a data storage area. The program storage area may store the operating system, applications required for at least one function (such as sound playback, image playback, etc.), etc.; the data storage area may store data created based on the use of the mobile phone (such as audio data, phonebook, etc.). Furthermore, the memory 409 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0152] The processor 410 is the control center of the electronic device. It connects various parts of the electronic device via various interfaces and lines. By running or executing software programs and / or modules stored in the memory 409, and by calling data stored in the memory 409, it performs various functions and processes data, thereby providing overall monitoring of the electronic device. The processor 410 may include one or more processing units; preferably, the processor 410 may integrate an application processor and a modem processor. The application processor mainly handles the operating system, user interface, and applications, while the modem processor mainly handles wireless communication. It is understood that the modem processor may not be integrated into the processor 410.

[0153] The electronic device 400 may also include a power supply 411 (such as a battery) for supplying power to various components. Preferably, the power supply 411 can be logically connected to the processor 410 through a power management system, thereby enabling functions such as managing charging, discharging, and power consumption through the power management system.

[0154] In addition, the electronic device 400 includes some functional modules not shown, which will not be described in detail here.

[0155] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0156] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0157] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of the present invention.

[0158] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed in this invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0159] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0160] In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0161] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0162] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0163] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, ROM, RAM, magnetic disks, or optical disks.

[0164] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. An electrical fault prediction method based on a neuron model, characterized in that, include: Construct a neuron model and a weighted regression model, wherein the neuron model includes at least an input layer, a hidden layer, and an output layer; Determine the business weight identifier and the neural node identifier of the neuron model; wherein, the business weight identifier and the neural node identifier are the data brought in when the data of the input layer is processed into the hidden layer; Obtain fault diagnosis sample data of a specified electrical appliance, and train the neuron model based on the fault diagnosis sample data of the specified electrical appliance, the business weight identifier, and the neural node identifier; The weighted regression model is used to determine the fitted values ​​of the input layer and the output layer of the trained neuron model; When the fitted value meets the preset threshold, the trained neuron model is determined as the trained neuron model, so as to predict the fault of the specified electrical appliance based on the trained neuron model. The hidden layer includes multiple neurons. After acquiring fault diagnosis sample data of a specified appliance and training the neuron model based on the fault diagnosis sample data of the specified appliance, the business weight identifier, and the neural node identifier, the method further includes: Determine the difference between the number of neurons currently allocated in the hidden layer and the number of neurons allocated in the previous time; when the difference exceeds a preset difference threshold, adjust the business weight identifier and the neural node identifier.

2. The method according to claim 1, characterized in that, After determining the fitted values ​​of the input layer and the output layer of the trained neuron model using the weighted regression model, the method further includes: When the fitted value does not meet the preset threshold, the business weight identifier and the neural node identifier are adjusted, and the process returns to the step of using the weight regression model to determine the fitted values ​​of the input layer and the output layer of the trained neuron model.

3. The method according to claim 1, characterized in that, The process of determining the business weight identifier and neural node identifier of the neuron model includes: The business weight identifier is determined based on the historical database topology of the neuron model, and the neural node identifier is determined based on the historical data of the same type in the neuron model. When it is impossible to determine the business weight identifier based on the historical database topology of the neuron model, and it is impossible to determine the neural node identifier based on the historical data of the same type of neuron model, the business weight identifier and the neural node identifier are randomly assigned.

4. The method according to claim 1, characterized in that, The weighted regression model includes a ridge regression method model, and the formula for the ridge regression method model is: Among them, the Indicates input data, the Indicates the output data, the This represents the training result of the neuron model, the To fit the hyperparameters; The = I, Among them, the Represents the identity matrix, the It is a weight constant, the stated It is the identity matrix The weights, the Represented as Seek under certain conditions .

5. An electrical fault prediction device based on a neuron model, characterized in that, include: The model building module is used to build a neuron model and a weighted regression model, wherein the neuron model includes at least an input layer, a hidden layer and an output layer; The identifier determination module is used to determine the business weight identifier and the neural node identifier of the neuron model; wherein, the business weight identifier and the neural node identifier are the data brought in when the data of the input layer is processed into the hidden layer; The model training module is used to acquire fault diagnosis sample data of a specified electrical appliance, and train the neuron model based on the fault diagnosis sample data of the specified electrical appliance, the business weight identifier, and the neural node identifier. The fitting value acquisition module is used to determine the fitting values ​​of the input layer and the output layer of the trained neuron model using the weighted regression model; The training determination module is used to determine the trained neuron model as the trained neuron model when the fitted value meets a preset threshold, so as to predict the fault of the specified electrical appliance based on the trained neuron model. The hidden layer includes multiple neurons. After acquiring fault diagnosis sample data of a specified appliance and training the neuron model based on the fault diagnosis sample data of the specified appliance, the business weight identifier, and the neural node identifier, the device further includes: Determine the difference between the number of neurons currently allocated in the hidden layer and the number of neurons allocated in the previous time; when the difference exceeds a preset difference threshold, adjust the business weight identifier and the neural node identifier.

6. The apparatus according to claim 5, characterized in that, The device further includes a return execution module for: When the fitted value does not meet the preset threshold, the business weight identifier and the neural node identifier are adjusted, and the process returns to the step of using the weight regression model to determine the fitted values ​​of the input layer and the output layer of the trained neuron model.

7. The apparatus according to claim 5, characterized in that, The identifier determination module is specifically used for: The business weight identifier is determined based on the historical database topology of the neuron model, and the neural node identifier is determined based on the historical data of the same type in the neuron model. When it is impossible to determine the business weight identifier based on the historical database topology of the neuron model, and it is impossible to determine the neural node identifier based on the historical data of the same type of neuron model, the business weight identifier and the neural node identifier are randomly assigned.

8. An electronic device, characterized in that, It includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; The memory is used to store computer programs; When the processor executes a program stored in the memory, it implements the method as described in any one of claims 1-4.

9. A computer-readable storage medium having instructions stored thereon that, when executed by one or more processors, cause the processors to perform the method as described in any one of claims 1-4.

Images