Power cable heating state monitoring and fault prediction method based on artificial intelligence

Abstract

The invention discloses a power cable heating state monitoring and fault prediction method based on artificial intelligence, and relates to the technical field of cables, and the method comprises the steps: collecting power cable data in an operation state, and marking the collected power cable data; a training framework of the state monitoring and fault prediction model is a meta learning and transfer learning fusion framework; a neural network based on feature dynamic influence is adopted; and an auto-encoder based on feature path decoupling is adopted. According to the method, the feature independence is optimized through a feature path decoupling mechanism, the independence among the power cable data features is evaluated and optimized through a regularization loss function, in addition, linear dependence among the power cable data features is calculated by adopting covariance, the expression ability of the independent power cable data features is enhanced through a penalty mechanism, and the power cable data features are more accurately extracted. The processing capability of the model to the power cable data is improved, and the model can still keep accurate expression of the power cable data structure even under the condition that the power cable data features are missing.

Description

Technical Field

[0001] The present invention relates to the field of cable technology, and in particular to a method for monitoring heating status and predicting faults of power cables based on artificial intelligence. Background Art

[0002] In the task of heating status monitoring and fault prediction of power cables, real-time processing of high-dimensional sensor data and effective feature extraction are the key to achieving accurate monitoring and prediction. The large amount of sensor data generated during the operation of power cables is multidimensional and dynamic, including complex information such as temperature, load current, and ambient humidity. These data are often highly correlated and redundant, which poses challenges to fault identification and status assessment. At the same time, in practical applications, sensors may lose some data due to environmental or equipment problems, which requires monitoring and prediction models to have high robustness and be able to accurately capture key information under incomplete data conditions. In addition, the operating environment of the power system is complex, and the distribution of data features will change with time and working conditions. Existing technologies have significant deficiencies in real-time adjustment of feature importance and optimization of fault prediction accuracy in dynamic environments.

[0003] In the prior art, the Chinese invention patent with application number CN118735141A proposes a cable operation and maintenance strategy optimization method based on machine learning, which belongs to the field of cable operation and maintenance technology, and specifically includes: setting the target area and inspection cycle and obtaining the inspection data of all cables; dividing the inspection cycle into n inspection cycles, selecting any cable and obtaining the number of times the cable fails in any inspection cycle and calculating the reference value J' of the cable; selecting any cable and calculating the time interval between adjacent faults, determining the number of inspection cycles in which the number of times any cable fails is greater than or equal to the reference value and marking it as the reference number; calculating the inspection score of any cable; correcting the inspection score of the cable according to the detection data, and taking the cable with the inspection score greater than or equal to the preset risk threshold as the target cable that needs to be inspected. The Chinese invention patent with publication number CN118708924A proposes a robot cable electrical performance testing method, device and system. The method includes: confirming the data change stability coefficient of each sampling time point based on each time period and a preset fixed value; obtaining the influence degree coefficient based on the data change stability coefficient of each sampling time point and the total number of sampling time points in each electrical performance factor, the total number of sampling time points and the preset response sensitivity threshold of each electrical performance factor; obtaining the efficiency influence trend coefficient of signal transmission of each electrical performance factor based on the monitoring data, time difference and influence degree coefficient of each sampling time point; testing the electrical performance of the robot cable based on the efficiency influence trend coefficient of signal transmission of each electrical performance factor to obtain the test results. The Chinese invention patent with publication number CN118228101A proposes a cable anomaly detection method and device, which includes: obtaining historical abnormal monitoring data with cable anomalies and historical normal monitoring data with normal cables, comparing and analyzing them, determining abnormal parameters according to the comparative analysis results, screening out historical abnormal parameter data corresponding to the abnormal parameters from the historical abnormal monitoring data, analyzing them, extracting feature data from the historical abnormal parameter data, and preprocessing them, building an initial model for cable anomaly detection based on multiple groups of preprocessed feature data and a preset machine algorithm, and training and testing the initial model for cable anomaly detection to obtain a cable anomaly detection model; judging whether the cable is abnormal by the cable anomaly detection model.

[0004] The above technical solution has the following problems: 1. In the task of monitoring the heating state of power cables and predicting faults, the existing feature extraction technology fails to effectively solve the redundancy and coupling problems between features, resulting in the power cable heating state monitoring and fault prediction model being difficult to maintain the integrity of the power cable data structure when features are missing or redundant features exist. 2. In the task of monitoring the heating state of power cables and predicting faults, the existing model lacks the ability to adjust feature weights in real time, resulting in the power cable heating state monitoring and fault prediction model being unstable when the power cable data distribution changes. 3. In the task of monitoring the heating state of power cables and predicting faults, the current autoencoder method usually only focuses on the optimization of reconstruction errors, ignoring the problem of maximizing information gain in the process of power cable data compression, which easily causes the loss of important information in the power cable data representation. Summary of the invention

[0005] The technical problem to be solved by the present invention is to overcome the shortcomings of the above-mentioned prior art and provide a method for monitoring the heating status and predicting faults of power cables based on artificial intelligence.

[0006] The technical solution adopted to solve the above technical problems is: a method for monitoring the heating state and predicting faults of power cables based on artificial intelligence, comprising the following steps: S1, collecting power cable data in operation, and marking the collected power cable data, the marked categories include: normal, warning, and fault; S2, the training framework of the condition monitoring and fault prediction model is a fusion framework of meta-learning and transfer learning; S3, uses a neural network based on the dynamic influence of features. The hidden layer of the neural network uses multi-scale feature fusion and high-order feature interaction strategies for dynamic calculation; S4 adopts an autoencoder based on feature path decoupling. It learns the compressed representation of power cable data through the autoencoder and optimizes the independence between features using the feature path decoupling mechanism. The feature path decoupling mechanism strengthens the independence between features through the regularized loss function, ensuring that even if some features are missing, the model can still capture the main structure of the power cable data.

[0007] Furthermore, the power cable data in S1 includes the real-time current value Ra of the cable, the temperature value Da of the cable, the ambient temperature Ca, the ambient humidity Fa, the voltage value Ga of the cable, the wind speed Va, the solar radiation intensity Ma, the historical load record Ba of the cable, the aging degree of the insulating material Na, and the connection status Ka of the cable connector.

[0008] Furthermore, the training process of the fusion framework in S2 includes a pre-training stage and a fine-tuning stage; In the pre-training stage, the power cable data samples are input into the feature extraction modules of transfer learning and meta-learning, wherein the power cable data samples with batches as training units are input into the transfer learning feature extraction module, and the power cable data samples with tasks as training units are input into the meta-learning feature extraction module; In the fine-tuning stage, the training set of the target domain power cable data is divided into batches and tasks as basic units. The training set of the target domain power cable data is the set of labeled power cable data in the target domain power cable data. The pre-trained condition monitoring and fault prediction models of meta-learning and transfer learning are used to classify the target domain power cable data, and the scores of the classification results of the two condition monitoring and fault prediction models are fused as the final classification results of the fusion framework.

[0009] Furthermore, the meta-learning adopts a model weak correlation meta-learning strategy, and the learning goal of the model weak correlation meta-learning strategy is to quickly learn and adapt to the optimal parameters required for new tasks through a parameter optimization algorithm, including a two-layer loop mode of inner loop and outer loop; The inner loop uses a basic learner to extract features of a specific task, and the basic learner is a neural network model; the outer loop uses a meta-learning strategy to update the initialization parameters of the basic learner through gradient descent.

[0010] Furthermore, the training process of the neural network algorithm based on the dynamic influence of features in S3 includes the following steps: S301, initialize the structure and parameters of the neural network to adapt to the distribution characteristics of different power cable data sets, expressed as: , In the formula, For the neural network The weight parameters of the layer, The mean is , the variance is Initialize with a normal distribution, is the mean of the neural network parameter initialization, is the variance of the neural network parameter initialization, An orthogonal perturbation matrix aligned with the input dimension ensures that the initial parameters of the neural network are highly sensitive to the input features; In order to optimize the sparseness of the neural network, the pruning threshold of the neural network is estimated according to the complexity of the input power cable data during initialization, which is expressed as: , In the formula, is the pruning threshold of the neural network, is the pruning adjustment coefficient of the neural network, is the characteristic variance of the input power cable data, is the variance function, is the power cable data feature input to the neural network; S302: For the power cable data input into the neural network, a feature redundancy suppression method is adopted to weaken the influence of redundant features through dynamic adjustment based on feature correlation weight parameters. The calculation method of the correlation weight parameters is expressed as: , In the formula, For the The first The correlation coefficient of the features, Features and The covariance of is the first input to the neural network Features, is the first input to the neural network Features, and Characteristics and The variance of The index of the input feature for the first neural network, The index of the input feature for the second neural network; According to the feature correlation weight parameter, the weight of the input feature is dynamically adjusted, which is expressed as: , In the formula, is the adjusted input to the neural network Features S303. The hidden layer of the neural network adopts a multi-scale feature fusion and high-order feature interaction strategy to achieve progressive abstraction of features from low-order to high-order and enhance the interactive representation capability between features. The multi-scale feature fusion method is expressed as: , In the formula, is the first input to the neural network The fusion features of the layers, is the index of the first neural network layer, is the number of multi-scales, s is the index of the scale, is the fusion weight of the s-th scale, is the training parameter, obtained through dynamic learning, is the nonlinear activation function of the s-th scale, The neural network The weight parameter of the s-th scale of the layer, is the fusion feature input to the l-1th layer of the neural network, The neural network Bias parameter of the sth scale of the layer; The nonlinear activation function used by the neural network can enhance the expression ability of nonlinear regions and ensure the smooth propagation of gradients in positive and negative regions through dynamic adjustment. The calculation method is expressed as: , In the formula, The neural network The nonlinear activation function of the layer, is the input of the nonlinear activation function, is the activation function of the rectified linear unit, is the first dynamic adjustment factor, is the hyperbolic tangent activation function, is the second dynamic regulatory factor; S304, dynamically adjust the gradient update amplitude by calculating the importance coefficient of each layer feature of the neural network, and the calculation method is expressed as: , In the formula, For the neural network The weight update increment of the layer, is the learning rate of the neural network, For the neural network The entropy of the layer gradient, is the symbol of partial derivative, is the loss function of the neural network, specifically the cross entropy loss, which is obtained by calculating the category probability of the output features of the neural network using the preset Softmax function and then calculating it through the cross entropy loss function. The neural network The weight parameters of the layer, For the neural network Importance factor of layer features; The feature importance factor dynamically measures the importance of the feature to the global output to ensure a more reasonable gradient distribution. The calculation method is expressed as: , In the formula, is the variance function, is the first input to the neural network The fusion features of the layers, is the total number of layers in the neural network, is the index of the second neural network layer; The learning rate of the neural network is adjusted based on the rate of change of the gradient, and the calculation method is expressed as: , In the formula, For the neural network The base learning rate of the layer, For the neural network The learning rate adjustment coefficient of the layer indicates the impact of gradient changes on the learning rate. for norm, is the updated gradient of the loss function in this iteration and the previous iteration; During the training process of each layer, the entropy value of the neural network gradient is calculated to ensure that the learning speed of each layer is coordinated. The calculation method is expressed as: , In the formula, The neural network The number of neurons in the layer, For the neural network Tier The gradient of a neuron, For the neural network The index of the layer neuron, is a logarithmic function; Update the neural network weight parameters, and the update method is expressed as: , In the formula, is the parameter update operation, For the neural network Regularization factor of the layer; In order to avoid overfitting, the regularization factor of the neural network is related to the entropy value of the gradient. The regularization factor decreases as the gradient increases, thereby reducing the model's sensitivity to overfitting and effectively adjusting the learning progress of each layer in the network. The calculation method is expressed as: , In the formula, The neural network The hyperparameters of the layer’s regularization factor affect the degree of influence. is the weight update gradient of the neural network, for norm; S305, repeat the above steps until the preset stop iteration condition is met. After the neural network model training is completed, the output feature of the neural network is the feature vector after feature extraction. The feature vector is passed through a preset Softmax function to obtain the prediction score of the power cable data.

[0011] Furthermore, the training process of the autoencoder algorithm based on feature path decoupling in S4 includes the following steps: S401, initialize the encoder and decoder of the autoencoder. The encoder is responsible for mapping the high-dimensional input power cable data to a low-dimensional space, and the decoder attempts to reconstruct the original power cable data from the low-dimensional representation. The initialization method is expressed as: , In the formula, and The encoder and decoder are The weight matrix of the layer, and The encoder and decoder are The bias vector of the layer, For encoder The dimension of the layer, For encoder The dimension of the layer, and to generate normally distributed random numbers and zero vectors, respectively. is the layer index of the encoder and decoder; S402. During the encoding process, the dependency path of the input features is decoupled by the regularization method. The loss function calculation method of the autoencoder is expressed as: , In the formula, is the total loss function of the autoencoder, including reconstruction error and regularization term, x represents the independent variable of the function, is the encoder function, is the decoder function, is the regularization weight, is the decoupling regularization function; The role of the decoupling regularization function is to evaluate the independence of features, and the calculation method is expressed as: , Where z is the input of the decoupling regularization function, The index of the input feature for the first decoupling function, The index of the input feature for the second decoupling function, is the tth iteration The weight of the feature, Representation characteristics and The covariance of , which measures the linear dependence between features, The first input of the decoupling function Features, The first input of the decoupling function Features, is the decoupling function variance weight factor, The index of the input feature for the third decoupling function, Features The variance of , enables the decoupling function to improve the information retention ability by suppressing small variance features, The first input of the decoupling function Features S403, through the dynamic feature flow embedding mechanism, the model dynamically adjusts the weight of each feature in the hidden layer according to the reconstruction error and feature importance to adapt to the power cable data characteristics and task requirements. The adjustment method is expressed as: , In the formula, is the t+1th iteration The weight of the feature, is the tth iteration The weight of the feature, is the dynamic feature flow embedding learning rate, is the gradient of the loss function with respect to the feature weight, Embedding tuning parameters for dynamic feature flow, For the The information gain of each feature; S404, the training process adopts a cyclic process of feedforward, feedback and parameter update, and minimizes the loss function through the gradient descent method. The parameter update method of the encoder is expressed as: , In the formula, is the parameter update operation, and are the loss functions of the autoencoder for the Gradients of layer weights and biases; S405, repeat the above iterative process, measure the model performance through comprehensive evaluation indicators, and optimize the reconstruction error and feature information. The calculation method of the evaluation indicator is expressed as: , In the formula, is the training evaluation index of the autoencoder, is the first adjustment weight of the autoencoder, is the second adjustment weight of the autoencoder, For the The information gain of each feature; Repeat iteration S402-S405 until the evaluation index is less than the preset threshold, then stop the iteration, which means the model training is completed. After the autoencoder model training is completed, the output feature of the encoder is the low-dimensional feature vector after feature extraction. The low-dimensional feature vector is passed through the preset Softmax function to obtain the prediction score of the power cable data.

[0012] The beneficial effects of the present invention are as follows: (1) In the tasks of power cable heating status monitoring and fault prediction, the present invention optimizes feature independence through a feature path decoupling mechanism, and evaluates and optimizes the independence between power cable data features through a regularized loss function. In addition, the covariance is used to calculate the linear dependency between power cable data features, and the penalty mechanism is used to enhance the expression ability of independent power cable data features, thereby improving the model's ability to process power cable data. Even when the power cable data features are missing, the model can still maintain an accurate expression of the power cable data structure.

[0013] (2) In the task of power cable heating state monitoring and fault prediction, the present invention utilizes a dynamic feature stream embedding mechanism to adapt to the data characteristics of power cables, dynamically adjusts the feature weights of the hidden layer according to the importance of the power cable data features, and optimizes the feature representation in real time through gradient updating. In addition, combined with reconstruction error and information gain, dynamic feature stream embedding enables the power cable heating state monitoring and fault prediction model to adaptively adjust the feature weight distribution under different data distributions and task requirements.

[0014] (3) In the task of power cable heating state monitoring and fault prediction, the present invention proposes a comprehensive evaluation index combining reconstruction error and feature information gain, guides the training of power cable heating state monitoring and fault prediction model in a multi-objective optimization manner, and balances the reconstruction ability and information retention ability of the model by adjusting parameters, effectively avoiding the problem of power cable data feature loss during information compression. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a principle block diagram of the fusion framework in the present invention. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0017] The method for monitoring heating status and predicting faults of power cables based on artificial intelligence includes the following steps: S1, collects the data of the power cables in operation, and marks the collected data. The marked categories include: Normal: the cable is operating in a normal state, without overload or abnormal heating. Warning: the cable may have slight heating or is about to enter an overload state. Fault: the cable has severe heating or has already failed.

[0018] The power cable data includes the real-time current value Ra of the cable, which reflects the cable load; the temperature value Da of the cable, which affects the safe operation of the cable; the ambient temperature Ca, which affects the cable temperature; the ambient humidity Fa, which is an important factor affecting the cable performance; the voltage value Ga of the cable, which reflects the power transmission efficiency; the wind speed Va, which affects the heat dissipation of the cable; the solar radiation intensity Ma, which has a direct impact on the cable temperature; the historical load record Ba of the cable, which is used for trend analysis; the aging degree of the insulating material Na, which measures the health status of the cable; and the connection status of the cable joint Ka, which causes local heating of the cable.

[0019] S2, the training framework of the condition monitoring and fault prediction model is a fusion framework of meta-learning and transfer learning, such as Figure 1 shown.

[0020] The training process of the fusion framework includes the pre-training phase and the fine-tuning phase: In the pre-training stage, the power cable data samples are input into the feature extraction modules of transfer learning and meta-learning, wherein the power cable data samples with batches as training units are input into the transfer learning feature extraction module, and the power cable data samples with tasks as training units are input into the meta-learning feature extraction module.

[0021] In the fine-tuning stage, the training set of the target domain power cable data is divided into batches and tasks as basic units. The training set of the target domain power cable data is the set of labeled power cable data in the target domain power cable data. The pre-trained condition monitoring and fault prediction models of meta-learning and transfer learning are used to classify the target domain power cable data, and the scores of the classification results of the two condition monitoring and fault prediction models are fused as the final classification results of the fusion framework.

[0022] Meta-learning learns meta-knowledge of different tasks through a multi-task learning paradigm, and uses meta-knowledge to quickly learn in new tasks, so that it can achieve good learning results with a small number of power cable data samples. Transfer learning relies on the knowledge transfer from the source domain power cable data to the target domain power cable data to solve the problem of small samples and cross-domain learning of power cable data. It requires a large amount of labeled power cable data to pre-train the state monitoring and fault prediction model.

[0023] Meta-learning adopts the model weak correlation meta-learning strategy. The learning goal of the model weak correlation meta-learning strategy is to quickly learn and adapt to the optimal parameters required for new tasks through parameter optimization algorithms, including two-layer loop modes of inner loop and outer loop. The inner loop uses the basic learner to extract the features of specific tasks. The basic learner is a neural network model. The outer loop uses the meta-learning strategy to update the initialization parameters of the basic learner through gradient descent.

[0024] The meta-learning strategy maps each sample in the support set and each sample in the query set to a high-dimensional feature space as the input of the neural network model. During the training process, several categories are randomly selected for condition monitoring and fault prediction model training. After completing the task learning of all batches, the optimal initialization parameters are obtained. When encountering a new task, the optimal initialization parameters are fine-tuned using the support set of the new task to obtain the most suitable parameters for the new task.

[0025] S3 adopts a neural network based on the dynamic influence of features. The hidden layer of the neural network adopts a multi-scale feature fusion and high-order feature interaction strategy to achieve progressive abstraction of features from low-order to high-order. It can not only effectively capture the diversity of high-dimensional power cable data, but also improve the training efficiency and reasoning speed of the neural network, while enhancing the generalization performance of the model and its adaptability to complex power cable data.

[0026] The training process of the neural network algorithm based on feature dynamic influence includes the following steps: S301, initialize the structure and parameters of the neural network to adapt to the distribution characteristics of different power cable data sets, expressed as: , In the formula, For the neural network The weight parameters of the layer, The mean is , the variance is Initialize with a normal distribution, is the mean of the neural network parameter initialization, is the variance of the neural network parameter initialization, It is an orthogonal perturbation matrix aligned with the input dimension, ensuring that the initial parameters of the neural network are highly sensitive to the input features.

[0027] In order to optimize the sparseness of the neural network, the pruning threshold of the neural network is estimated according to the complexity of the input power cable data during initialization, which is expressed as: , In the formula, is the pruning threshold of the neural network, is the pruning adjustment coefficient of the neural network, Set to 0.2, is the characteristic variance of the input power cable data, is the variance function, The power cable data features input to the neural network.

[0028] When the pruning threshold of the neural network is greater than the preset threshold, a preset proportion of neurons are randomly inactivated to reduce redundant connections in the initialization phase while retaining the core feature pathways.

[0029] S302: For the power cable data input into the neural network, a feature redundancy suppression method is adopted to weaken the influence of redundant features through dynamic adjustment based on feature correlation weight parameters. The calculation method of the correlation weight parameters is expressed as: , In the formula, For the The first The correlation coefficient of the features, Features and The covariance of is the first input to the neural network Features, is the first input to the neural network Features, and Characteristics and The variance of The index of the input feature for the first neural network, The indices of the input features for the second neural network.

[0030] According to the feature correlation weight parameter, the weight of the input feature is dynamically adjusted, which is expressed as: , In the formula, is the adjusted input to the neural network Features.

[0031] S303. The hidden layer of the neural network adopts a multi-scale feature fusion and high-order feature interaction strategy to achieve progressive abstraction of features from low-order to high-order and enhance the interactive representation capability between features. The multi-scale feature fusion method is expressed as: , In the formula, is the first input to the neural network The fusion features of the layers, is the index of the first neural network layer, is the number of multi-scales, s is the index of the scale, is the fusion weight of the s-th scale, is the training parameter, obtained through dynamic learning, is the nonlinear activation function of the s-th scale, The neural network The weight parameter of the s-th scale of the layer, is the fusion feature input to the l-1th layer of the neural network, The neural network Bias parameter of the s-th scale of the layer.

[0032] The nonlinear activation function used by the neural network can improve the expression ability of the nonlinear region, and at the same time ensure the smooth propagation of the gradient in the positive and negative regions through dynamic adjustment. The calculation method is expressed as: , In the formula, The neural network The nonlinear activation function of the layer, is the input of the nonlinear activation function, is the activation function of the rectified linear unit, is the first dynamic adjustment factor, is the hyperbolic tangent activation function, It is the second dynamic adjustment factor.

[0033] S304, dynamically adjust the gradient update amplitude by calculating the importance coefficient of each layer feature of the neural network, and the calculation method is expressed as: , In the formula, For the neural network The weight update increment of the layer, is the learning rate of the neural network, For the neural network The entropy of the layer gradient, is the symbol of partial derivative, is the loss function of the neural network, specifically the cross entropy loss, which is obtained by calculating the category probability of the output features of the neural network using the preset Softmax function and then calculating it through the cross entropy loss function. The neural network The weight parameters of the layer, For the neural network Importance factor of layer features.

[0034] The feature importance factor dynamically measures the importance of the feature to the global output to ensure a more reasonable gradient distribution. The calculation method is expressed as: , In the formula, is the variance function, is the first input to the neural network The fusion features of the layers, is the total number of layers in the neural network, is the index of the second neural network layer.

[0035] The learning rate of the neural network is adjusted based on the rate of change of the gradient. As the gradient changes, the learning rate is adjusted dynamically to avoid unstable learning caused by excessive update steps. The calculation method is expressed as: , In the formula, For the neural network The base learning rate of the layer, For the neural network The learning rate adjustment coefficient of the layer indicates the impact of gradient changes on the learning rate. for norm, is the updated gradient of the loss function in this iteration and the previous iteration.

[0036] During the training process of each layer, the entropy value of the neural network gradient is calculated to ensure that the learning speed of each layer is coordinated. The calculation method is expressed as: , In the formula, The neural network The number of neurons in the layer, For the neural network Tier The gradient of a neuron, For the neural network The index of the layer neuron, For the logarithmic function, the default base of the logarithmic function is set to 10.

[0037] Update the neural network weight parameters, and the update method is expressed as: , In the formula, is the parameter update operation, For the neural network Regularization factor for the layer.

[0038] In order to avoid overfitting, the regularization factor of the neural network is related to the entropy value of the gradient. The regularization factor decreases as the gradient increases, thereby reducing the model's sensitivity to overfitting and effectively adjusting the learning progress of each layer in the network. The calculation method is expressed as: , In the formula, The neural network The hyperparameters of the regularization factor of each layer are set to 0.1. is the weight update gradient of the neural network, is the L2 norm.

[0039] S305. Repeat iterations S302-S304 until a preset stop iteration condition is met. The preset stop iteration condition is reaching a preset maximum number of iterations. The preset maximum number of iterations is set to 1000 times, which means that the model training is completed. After the neural network model training is completed, the output feature of the neural network is the feature vector after feature extraction. The feature vector is passed through a preset Softmax function to obtain a prediction score for the power cable data.

[0040] In the fusion framework, unlike meta-learning based on the weak correlation strategy of the model, transfer learning can mine subtle features under the shallow features of power cable data. The advantage of transfer learning is that it can make full use of the source domain power cable data samples to pre-train the state monitoring and fault prediction models, and specifically uses the autoencoder model for migration.

[0041] S4 adopts an autoencoder based on feature path decoupling. It learns the compressed representation of power cable data through the autoencoder and optimizes the independence between features using the feature path decoupling mechanism. The feature path decoupling mechanism strengthens the independence between features through the regularized loss function, ensuring that even if some features are missing, the model can still capture the main structure of the power cable data.

[0042] The training process of the autoencoder algorithm based on feature path decoupling includes the following steps: S401, initialize the encoder and decoder of the autoencoder. The encoder is responsible for mapping the high-dimensional input power cable data to a low-dimensional space, and the decoder attempts to reconstruct the original power cable data from the low-dimensional representation. The initialization method is expressed as: , In the formula, and The encoder and decoder are The weight matrix of the layer, and The encoder and decoder are The bias vector of the layer, For encoder The dimension of the layer, For encoder The dimension of the layer, and to generate normally distributed random numbers and zero vectors, respectively. The encoder and decoder layer indices.

[0043] S402. During the encoding process, the dependency path of the input features is decoupled by a regularization method, and features that can independently express the structure of the power cable data are discovered and strengthened even when other feature information is removed. Feature path decoupling strengthens the independence between features through a regularized loss function, ensuring that even if some features are missing, the model can still capture the main structure of the power cable data. The loss function calculation method of the autoencoder is expressed as: , In the formula, is the total loss function of the autoencoder, including reconstruction error and regularization term. x represents the independent variable of the function. Here, only the power cable data is input into the autoencoder. is the encoder function, is the decoder function, is the regularization weight, Set to 0.45, is the decoupled regularization function.

[0044] The role of the decoupling regularization function is to evaluate the independence of features, and the calculation method is expressed as: , Where z is the input of the decoupling regularization function, The index of the input feature for the first decoupling function, The index of the input feature for the second decoupling function, is the tth iteration The weight of the feature, Representation characteristics and The covariance of , which measures the linear dependence between features, The first input of the decoupling function Features, The first input of the decoupling function Features, is the decoupling function variance weight factor, The index of the input feature for the third decoupling function, Features The variance of , enables the decoupling function to improve the information retention ability by suppressing small variance features, The first input of the decoupling function Features.

[0045] S403, through the dynamic feature flow embedding mechanism, the model dynamically adjusts the weight of each feature in the hidden layer according to the reconstruction error and feature importance to adapt to the power cable data characteristics and task requirements. The adjustment method is expressed as: , In the formula, is the t+1th iteration The weight of the feature, is the tth iteration The weight of the feature, is the dynamic feature flow embedding learning rate, Set to 0.01, is the gradient of the loss function with respect to the feature weight, Embedding tuning parameters for dynamic feature flow, Set to 0.4, For the The information gain of a feature.

[0046] S404, the training process adopts a cyclic process of feedforward, feedback and parameter update, and minimizes the loss function through the gradient descent method. The parameter update method of the encoder is expressed as: , In the formula, is the parameter update operation, and are the loss functions of the autoencoder for the Gradients of the layer's weights and biases.

[0047] S405, repeat the above iterative process, measure the model performance through comprehensive evaluation indicators, so as to optimize the reconstruction error and feature information at the same time, and the calculation method of the evaluation indicator is expressed as: , In the formula, is the training evaluation index of the autoencoder, is the first adjustment weight of the autoencoder, is the second adjustment weight of the autoencoder, For the The information gain of a feature.

[0048] Repeat iteration S402-S405 until the evaluation index is less than the preset threshold, then stop the iteration, which means the model training is completed. After the autoencoder model training is completed, the output feature of the encoder is the low-dimensional feature vector after feature extraction. The low-dimensional feature vector is passed through the preset Softmax function to obtain the prediction score of the power cable data.

[0049] After the pre-training, the parameters of the two feature extraction networks are fixed and transferred to the target domain power cable data for fine-tuning. The data samples of the target domain power cable data are divided into a support set and a query set according to the meta-learning strategy. The power cable data of the support set is used to fine-tune the state monitoring and fault prediction model of transfer learning. The fine-tuning strategy can adopt the gradient descent method.

[0050] After completing the fine-tuning of the condition monitoring and fault prediction model, the query set of the target domain power cable data is input into the neural network model and the autoencoder model respectively to obtain the corresponding prediction scores. The prediction scores are then normalized using the Softmax function. Finally, the two prediction scores are fused and output as the final prediction result. The classification category is the category corresponding to the maximum prediction score.

[0051] The above description is only a preferred embodiment of the present invention and is not intended to limit the protection scope of the present invention.

Claims

1. A method for monitoring heating status and predicting faults of power cables based on artificial intelligence, characterized in that: The following steps are involved: S1, collecting power cable data in operation, and marking the collected power cable data, the marked categories include: normal, warning, and fault; S2, the training framework of the condition monitoring and fault prediction model is a fusion framework of meta-learning and transfer learning; S3, uses a neural network based on the dynamic influence of features. The hidden layer of the neural network uses multi-scale feature fusion and high-order feature interaction strategies for dynamic calculation; S4 adopts an autoencoder based on feature path decoupling. It learns the compressed representation of power cable data through the autoencoder and optimizes the independence between features using the feature path decoupling mechanism. The feature path decoupling mechanism strengthens the independence between features through the regularized loss function, ensuring that even if some features are missing, the model can still capture the main structure of the power cable data.

2. The method for monitoring heating status and predicting faults of power cables based on artificial intelligence according to claim 1 is characterized in that: The power cable data in S1 include the real-time current value Ra of the cable, the temperature value Da of the cable, the ambient temperature Ca, the ambient humidity Fa, the voltage value Ga of the cable, the wind speed Va, the solar radiation intensity Ma, the historical load record Ba of the cable, the aging degree Na of the insulating material, and the connection status Ka of the cable joint.

3. The method for monitoring heating status and predicting faults of power cables based on artificial intelligence according to claim 1 is characterized in that: The training process of the fusion framework in S2 includes a pre-training stage and a fine-tuning stage; In the pre-training stage, the power cable data samples are input into the feature extraction modules of transfer learning and meta-learning, wherein the power cable data samples with batches as training units are input into the transfer learning feature extraction module, and the power cable data samples with tasks as training units are input into the meta-learning feature extraction module; In the fine-tuning stage, the training set of the target domain power cable data is divided into batches and tasks as basic units. The training set of the target domain power cable data is the set of labeled power cable data in the target domain power cable data. The pre-trained condition monitoring and fault prediction models of meta-learning and transfer learning are used to classify the target domain power cable data, and the scores of the classification results of the two condition monitoring and fault prediction models are fused as the final classification results of the fusion framework.

4. The method for monitoring heating status and predicting faults of power cables based on artificial intelligence according to claim 3 is characterized in that: The meta-learning adopts a model weak correlation meta-learning strategy. The learning goal of the model weak correlation meta-learning strategy is to quickly learn and adapt to the optimal parameters required for new tasks through a parameter optimization algorithm, including a two-layer loop mode of inner loop and outer loop; The inner loop uses a basic learner to extract features of a specific task, and the basic learner is a neural network model; the outer loop uses a meta-learning strategy to update the initialization parameters of the basic learner through gradient descent.

5. The method for monitoring heating status and predicting faults of power cables based on artificial intelligence according to claim 1, characterized in that: The training process of the neural network algorithm based on dynamic influence of features in S3 includes the following steps: S301, initialize the structure and parameters of the neural network to adapt to the distribution characteristics of different power cable data sets, expressed as: ， In the formula, For the neural network The weight parameters of the layer, The mean is , the variance is Initialize with a normal distribution, is the mean of the neural network parameter initialization, is the variance of the neural network parameter initialization, An orthogonal perturbation matrix aligned with the input dimension ensures that the initial parameters of the neural network are highly sensitive to the input features; In order to optimize the sparseness of the neural network, the pruning threshold of the neural network is estimated according to the complexity of the input power cable data during initialization, which is expressed as: ， In the formula, is the pruning threshold of the neural network, is the pruning adjustment coefficient of the neural network, is the characteristic variance of the input power cable data, is the variance function, is the power cable data feature input to the neural network; S302: For the power cable data input into the neural network, a feature redundancy suppression method is adopted to weaken the influence of redundant features through dynamic adjustment based on feature correlation weight parameters. The calculation method of the correlation weight parameters is expressed as: ， In the formula, For the The first The correlation coefficient of the features, Features and The covariance of is the first input to the neural network Features, is the first input to the neural network Features, and Characteristics and The variance of The index of the input feature for the first neural network, The index of the input feature for the second neural network; According to the feature correlation weight parameter, the weight of the input feature is dynamically adjusted, which is expressed as: ， In the formula, is the adjusted input to the neural network Features S303. The hidden layer of the neural network adopts a multi-scale feature fusion and high-order feature interaction strategy to achieve progressive abstraction of features from low-order to high-order and enhance the interactive representation capability between features. The multi-scale feature fusion method is expressed as: , In the formula, is the first input to the neural network The fusion features of the layers, is the index of the first neural network layer, S is the number of multi-scales, s is the index of the scale, is the fusion weight of the s-th scale, is the training parameter, obtained through dynamic learning, is the nonlinear activation function of the s-th scale, The neural network The weight parameter of the s-th scale of the layer, is the fusion feature input to the l-1th layer of the neural network, The neural network Bias parameter of the sth scale of the layer; The nonlinear activation function used by the neural network can enhance the expression ability of nonlinear regions and ensure the smooth propagation of gradients in positive and negative regions through dynamic adjustment. The calculation method is expressed as: ， In the formula, The neural network The nonlinear activation function of the layer, is the input of the nonlinear activation function, is the activation function of the rectified linear unit, is the first dynamic adjustment factor, is the hyperbolic tangent activation function, is the second dynamic regulatory factor; S304, dynamically adjust the gradient update amplitude by calculating the importance coefficient of each layer feature of the neural network, and the calculation method is expressed as: , In the formula, For the neural network The weight update increment of the layer, is the learning rate of the neural network, For the neural network The entropy of the layer gradient, is the symbol of partial derivative, is the loss function of the neural network, specifically the cross entropy loss, which is obtained by calculating the category probability of the output features of the neural network using the preset Softmax function and then calculating it through the cross entropy loss function. The neural network The weight parameters of the layer, For the neural network Importance factor of layer features; The feature importance factor dynamically measures the importance of the feature to the global output to ensure a more reasonable gradient distribution. The calculation method is expressed as: ， In the formula, is the variance function, is the first input to the neural network The fusion features of the layers, is the total number of layers in the neural network, is the index of the second neural network layer; The learning rate of the neural network is adjusted based on the rate of change of the gradient, and the calculation method is expressed as: , In the formula, For the neural network The base learning rate of the layer, For the neural network The learning rate adjustment coefficient of the layer indicates the impact of gradient changes on the learning rate. for norm, is the updated gradient of the loss function in this iteration and the previous iteration; During the training process of each layer, the entropy value of the neural network gradient is calculated to ensure that the learning speed of each layer is coordinated. The calculation method is expressed as: , In the formula, The neural network The number of neurons in the layer, For the neural network Tier The gradient of a neuron, For the neural network The index of the layer neuron, is a logarithmic function; Update the neural network weight parameters, and the update method is expressed as: ， In the formula, is the parameter update operation, For the neural network Regularization factor of the layer; In order to avoid overfitting, the regularization factor of the neural network is related to the entropy value of the gradient. The regularization factor decreases as the gradient increases, thereby reducing the model's sensitivity to overfitting and effectively adjusting the learning progress of each layer in the network. The calculation method is expressed as: ， In the formula, The neural network The hyperparameters of the layer’s regularization factor affect the degree of influence. is the weight update gradient of the neural network, for norm; S305, repeat the above steps until the preset stop iteration condition is met. After the neural network model training is completed, the output feature of the neural network is the feature vector after feature extraction. The feature vector is passed through a preset Softmax function to obtain the prediction score of the power cable data.

6. The method for monitoring heating status and predicting faults of power cables based on artificial intelligence according to claim 1, characterized in that: The training process of the autoencoder algorithm based on feature path decoupling in S4 includes the following steps: S401, initialize the encoder and decoder of the autoencoder. The encoder is responsible for mapping the high-dimensional input power cable data to a low-dimensional space, and the decoder attempts to reconstruct the original power cable data from the low-dimensional representation. The initialization method is expressed as: ， In the formula, and The encoder and decoder are The weight matrix of the layer, and The encoder and decoder are The bias vector of the layer, For encoder The dimension of the layer, For encoder The dimension of the layer, and to generate normally distributed random numbers and zero vectors, respectively. is the layer index of the encoder and decoder; S402. During the encoding process, the dependency path of the input features is decoupled by the regularization method. The loss function calculation method of the autoencoder is expressed as: ， In the formula, is the total loss function of the autoencoder, including reconstruction error and regularization term, x represents the independent variable of the function, is the encoder function, is the decoder function, is the regularization weight, is the decoupling regularization function; The role of the decoupling regularization function is to evaluate the independence of features, and the calculation method is expressed as: ， Where z is the input of the decoupling regularization function, The index of the input feature for the first decoupling function, The index of the input feature for the second decoupling function, is the tth iteration The weight of the feature, Representation characteristics and The covariance of , which measures the linear dependence between features, The first input of the decoupling function Features, The first input of the decoupling function Features, is the decoupling function variance weight factor, The index of the input feature for the third decoupling function, Features The variance of , enables the decoupling function to improve the information retention ability by suppressing small variance features, The first input of the decoupling function Features S403, through the dynamic feature flow embedding mechanism, the model dynamically adjusts the weight of each feature in the hidden layer according to the reconstruction error and feature importance to adapt to the power cable data characteristics and task requirements. The adjustment method is expressed as: ， In the formula, is the t+1th iteration The weight of the feature, is the tth iteration The weight of the feature, is the dynamic feature flow embedding learning rate, is the gradient of the loss function with respect to the feature weight, Embedding tuning parameters for dynamic feature flow, For the The information gain of each feature; S404, the training process adopts a cyclic process of feedforward, feedback and parameter update, and minimizes the loss function through the gradient descent method. The parameter update method of the encoder is expressed as: ， In the formula, is the parameter update operation, and are the loss functions of the autoencoder for the Gradients of layer weights and biases; S405, repeat the above iterative process, measure the model performance through comprehensive evaluation indicators, and optimize the reconstruction error and feature information. The calculation method of the evaluation indicator is expressed as: ， In the formula, is the training evaluation index of the autoencoder, is the first adjustment weight of the autoencoder, is the second adjustment weight of the autoencoder, For the The information gain of each feature; Repeat iteration S402-S405 until the evaluation index is less than the preset threshold, then stop the iteration, which means the model training is completed. After the autoencoder model training is completed, the output feature of the encoder is the low-dimensional feature vector after feature extraction. The low-dimensional feature vector is passed through the preset Softmax function to obtain the prediction score of the power cable data.

Images