Cable core temperature prediction method and device and storage medium
The LSTM network is optimized by combining the finite element method and the Laplace fraction algorithm with the genetic algorithm, and the error and efficiency problems in the calculation of cable core temperature are solved, and the cable core temperature prediction with high accuracy and interpretability is achieved.
Patent Information
- Application Number
- CN202411093088.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-09
- Publication Date
- 2024-12-06
AI Technical Summary
The prior art has large errors in the calculation of cable core temperature. Especially in complex environments, the finite element method has a long calculation time, making it difficult to meet the requirements of high-quality development of the power grid for cable reliability and stability.
The cable core temperature prediction method based on the finite element method and Laplace fraction is adopted. By obtaining the cable parameter information, a finite element model is established, and the finite element temperature field analysis is performed. The combination of the Laplace fraction algorithm is important to characterize the temperature timing data. and genetic algorithms are used to optimize long-term short-term memory network (LSTM) for training to build a cable temperature field prediction model.
Improve the accuracy and efficiency of cable core temperature prediction, enhance the interpretability of the model, dig out the nonlinear relationship between cable core temperature and predictive variables, and provide a construction of interpretable cable core temperature prediction theory. support.
Smart Images

Figure CN119089733A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of cable core temperature prediction, and in particular to a cable core temperature prediction method, device and storage medium. Background Art
[0002] The high-quality development of the power grid has put forward further requirements on the reliability and stability of power cables. If an accident occurs in the power cable, it will cause immeasurable economic losses. Therefore, it is particularly important to accurately determine the cable core temperature and then evaluate the operating status of the cable. The traditional cable temperature calculation methods are mainly divided into numerical calculation methods and analytical calculation methods. The analytical calculation method usually takes the steady-state temperature field theory as the theoretical basis to construct an equivalent thermal circuit model for the cable, but the calculation error is large in the actual complex environment. In contrast, the finite element method has the advantages of high simulation calculation accuracy and the ability to simulate complex environments. However, under the same cable model, the time required for finite element calculation has increased significantly. Summary of the invention
[0003] In order to solve at least one of the technical problems existing in the prior art to a certain extent, the object of the present invention is to provide a cable core temperature prediction method, device and storage medium based on finite element method and Laplace fraction.
[0004] The first technical solution adopted by the present invention is:
[0005] A method for predicting cable core temperature comprises the following steps:
[0006] Obtain parameter information of the target cable and establish a finite element model of the target cable;
[0007] Under different operating conditions, the finite element temperature field analysis and calculation are performed on the target cable finite element model to obtain the temperature field calculation results;
[0008] According to the parameter information and the temperature field calculation results of the cable finite element model, a cable core temperature time series data set is constructed;
[0009] The Laplace fractional algorithm is used to evaluate the importance of the input features in the cable core temperature time series data set, and the N input features with the largest contribution are selected as the final input features of the model; N is an integer greater than 1;
[0010] Based on the cable core temperature time series dataset after feature selection, the long short-term memory network (LSTM) optimized by genetic algorithm is trained to obtain the cable temperature field prediction model, which is used to predict the cable core temperature.
[0011] Furthermore, the parameter information of the target cable includes: structural parameters, thermal parameters, electromagnetic parameters, laying conditions and boundary conditions of the target cable;
[0012] The temperature field calculation results include: the internal cable core temperature data of the target cable and the corresponding external environment parameters, load parameters and time series;
[0013] The structural parameters, thermal parameters, electromagnetic parameters, laying conditions, external environmental parameters, load parameters and time series of the target cable are taken as the input characteristics of the model, and the cable core temperature is taken as the output variable of the model.
[0014] Furthermore, the parameter information of the target cable includes:
[0015] The cross-sectional area and thickness of the cable core, cable core shielding layer, insulation layer, metal shielding layer and outer sheath layer of the target cable; and,
[0016] Thermal conductivity of the cable core, cable core shielding layer, insulation layer, metal shielding layer and outer sheath layer of the target cable; and,
[0017] Soil thermal resistivity, soil temperature, ambient temperature, and convective heat transfer coefficient; and,
[0018] Cable core conductivity and relative magnetic permeability; and,
[0019] Cable loading parameters.
[0020] Furthermore, the Laplace fractional algorithm is combined to evaluate the feature importance of the input features in the cable core temperature time series data set, and the N input features with the largest contribution are selected as the final input features of the model, including:
[0021] Assume that the cable core temperature time series dataset has m samples, r features, and L r is the Laplace score of the rth feature, f ri is the rth feature of the i-th sample;
[0022] Construct a nearest neighbor graph G with m nodes, where the i-th node corresponds to sample x i ; If the sample x i and sample x j Very close in the sample space, that is, sample x i Located in sample x j Among the k nearest neighbors, or sample x j Located in sample x i Among the k nearest neighbors, there is an edge between nodes i and j in the nearest neighbor graph G; otherwise, nodes i and j are not connected;
[0023] Construct the weight matrix S, where The parameter t is a hyperparameter, ||·|| represents the Euclidean distance, and S ijRepresents each element in the weight matrix S, which can represent the topological structure of the data space;
[0024] The Laplace matrix L=DS is calculated based on the weight matrix S, where D is the diagonal matrix D=diag(S) generated by the weight matrix S, and the Laplace score of the rth feature is obtained as:
[0025]
[0026] In the formula, f r =[f r1 , f r2 , …, f rm ] T ,
[0027] Obtain N features as the final input features of the model based on the obtained Laplace scores;
[0028] The smaller the Laplace score is, the more local information this feature can retain and the better the discriminability is, which means that this feature is more important.
[0029] Furthermore, the long short-term memory network is composed of a forget gate, an input gate and an output gate;
[0030] The forget gate is used to determine whether information should be removed from the cell state. Its output is a number between 0 and 1, indicating the probability of information being forgotten. The calculation formula is:
[0031] f t =σ(W xf *x t +W lf *h t-1 +b f )
[0032] In the formula, f t Represents the output value of the forget gate at the current moment, x t Indicates the input at the current moment, h t-1 represents the hidden state at the previous moment, W xf and W hf Represents the input vector x t and the hidden state h t-1 and the weight matrix of the forget gate, b f represents the bias vector of the forget gate, σ represents the Sigmoid function;
[0033] The input gate is used to determine whether new information needs to be added to the cell state. The calculation formula is:
[0034]
[0035] In the formula, i t represents the input gate output value at the current moment, c t-1 Indicates the cell state at the previous moment, Represents a new candidate cell state, W xi and W hi Represents the input vector x t and the hidden state h t-1 and the weight matrix of the input gate, b i represents the bias vector of the input gate, and tanh represents the hyperbolic tangent function;
[0036] The output gate is used to determine which part of the cell state will be output to the hidden state, and the calculation formula is:
[0037] o t =σ(W xo *x t +W ho *h t-1 +b o )
[0038] h t =o t *tanh(c t )
[0039] In the formula, o t Indicates the output value of the output gate at the current moment, o t represents the hidden state at the current moment, W xo and W ho Represents the input vector x t and the hidden state h t-1 and the weight matrix of the output gate, b θ Represents the bias vector of the input gate.
[0040] Furthermore, the long short-term memory network is optimized in the following manner:
[0041] Genetic algorithms are used to optimize the number of layers of the long short-term memory network, the number of neurons in each layer, the number of hidden layers, and the learning rate. When the fitness function requirements are met, the optimization results, i.e., the optimal parameters of the model, are obtained.
[0042] The expression of the fitness function is:
[0043]
[0044] Where n represents the number of cable core temperature time series sample data in the test set, y i represents the real settlement data in the test set, represents the cable core temperature data predicted by the LSTM network, A represents a chromosome in the population, and R represents the preset hyperparameter value space.
[0045] Furthermore, the working mode of the genetic algorithm optimization is:
[0046] Initialize the population;
[0047] Repeatedly randomly select a number of individuals from the population to form a group, compare these individuals, and select the individual with the highest fitness as the parent individual for reproduction until m individuals are selected for subsequent genetic operations;
[0048] Crossover step: pair the selected m individuals in pairs without duplication, and generate a random number in the interval [0, 1] during the pairing process. If the random number is less than the preset crossover probability, a crossover operation is performed; otherwise, the two chromosomes remain the same;
[0049] Mutation step: For each of the m individuals that have undergone the crossover step, generate a random number in the interval [0,1]. If the generated random number is less than the preset mutation probability, a mutation operation is performed; otherwise, the individual remains the same; the mutated chromosome will be used as an individual in the offspring;
[0050] Repeat the crossover step and the mutation step until the preset number of genetic iterations is reached or the fitness function reaches the target value; select the chromosome with the best fitness from the population, which represents a set of optimal hyperparameter values, and use the optimal hyperparameter values to train the LSTM network, retaining the best weights obtained from the training.
[0051] Further, the cable temperature field prediction model is obtained to predict the cable core temperature, including:
[0052] Obtain the first i-1 data points of the cable core temperature time series data sequence, input them into the cable temperature field prediction model, predict and obtain the i-th data point, and add the obtained data point to the cable core temperature time series data sequence;
[0053] Remove the first data point in the cable core temperature time series data sequence, input the i-1 data point of the updated cable core temperature time series data sequence into the cable temperature field prediction model, and predict the i+1th data point;
[0054] The newly obtained data point is added to the cable core temperature time series data sequence, and the executed step of removing the first data point in the cable core temperature time series data sequence is returned.
[0055] The second technical solution adopted by the present invention is:
[0056] A cable core temperature prediction device, comprising:
[0057] at least one processor;
[0058] at least one memory for storing at least one program;
[0059] When the at least one program is executed by the at least one processor, the at least one processor implements the method described above.
[0060] The third technical solution adopted by the present invention is:
[0061] A computer-readable storage medium stores a program executable by a processor, wherein the program executable by the processor is used to execute the method described above when executed by the processor.
[0062] The fourth technical solution adopted by the present invention is:
[0063] A computer program product or a computer program includes computer instructions stored in a computer-readable storage medium. A processor of a computer device can read the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the above method.
[0064] The beneficial effects of the present invention are as follows: the present invention adopts the Laplace fractional algorithm to perform feature selection on the cable core temperature time series data, retains the features with high importance, and combines the genetic algorithm to optimize the parameters of the LSTM network, thereby realizing the cable core temperature prediction based on interpretable machine learning. The nonlinear relationship between the cable core temperature and the prediction variable is mined, the data utilization is enhanced, the accuracy of the model is improved, and theoretical support and new ideas are provided for constructing interpretable cable core temperature prediction. BRIEF DESCRIPTION OF THE DRAWINGS
[0065] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the embodiments of the present invention or the drawings of related technical solutions in the prior art are introduced below. It should be understood that the drawings introduced below are only for the convenience of clearly describing some embodiments of the technical solutions of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.
[0066] Figure 1 is a flow chart of the steps of a cable core temperature prediction method in an embodiment of the present invention;
[0067] Figure 2 is a schematic diagram of a physical model of cable laying in an embodiment of the present invention;
[0068] Figure 3 is a schematic diagram of finite element mesh generation of a single cable in an embodiment of the present invention;
[0069] Figure 4 is a schematic diagram of finite element mesh generation of a cable temperature field in an embodiment of the present invention;
[0070] Figure 5 It is an internal structure diagram of an LSTM network of a cable core temperature prediction method based on the finite element method in an embodiment of the present invention. DETAILED DESCRIPTION
[0071] The embodiments of the present invention are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and are not to be construed as limitations of the present invention. For the step numbers in the following embodiments, they are only provided for the convenience of explanation, and the order between the steps is not limited in any way, and the execution order of each step in the embodiment can be adaptively adjusted according to the understanding of those skilled in the art.
[0072] In the description of the present invention, it should be understood that descriptions involving orientations, such as up, down, front, back, left, right, etc., and orientations or positional relationships indicated are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present invention.
[0073] In the description of the present invention, "several" means one or more, "more" means more than two, "greater than", "less than", "exceed" etc. are understood as not including the number itself, and "above", "below", "within" etc. are understood as including the number itself. If there is a description of "first" or "second", it is only used for the purpose of distinguishing the technical features, and cannot be understood as indicating or implying the relative importance or implicitly indicating the number of the indicated technical features or implicitly indicating the order of the indicated technical features.
[0074] In the description of the present invention, unless otherwise clearly defined, terms such as setting, installing, connecting, etc. should be understood in a broad sense, and technicians in the relevant technical field can reasonably determine the specific meanings of the above terms in the present invention based on the specific content of the technical solution.
[0075] In view of the existing technical problems and taking into account the current application of machine learning methods in power grids, in order to improve the computing efficiency and model interpretability while ensuring the accuracy of cable temperature calculation, the present invention proposes a cable core temperature prediction method, device and storage medium based on finite element method and Laplace fraction.
[0076] like Figure 1 As shown, this embodiment provides a cable core temperature prediction method, which specifically includes the following steps:
[0077] S1. Obtain parameter information of the target cable and establish a finite element model of the target cable.
[0078] In some embodiments, the parameter information of the target cable includes: structural parameters, thermal parameters, electromagnetic parameters, laying conditions, and boundary conditions of the target cable.
[0079] As an optional implementation, see Figure 2 The parameter information specifically includes: the cross-sectional area and thickness of the target cable's core, core shielding layer, insulation layer, metal shielding layer, and outer sheath layer; the thermal conductivity of the target cable's core, core shielding layer, insulation layer, metal shielding layer, and outer sheath layer; soil thermal resistance, soil temperature, ambient temperature, and convection heat transfer coefficient; cable core conductivity and relative magnetic permeability; and cable load parameters.
[0080] S2. Under different operating conditions, the finite element temperature field analysis and calculation are performed on the target cable finite element model to obtain the temperature field calculation results.
[0081] In some embodiments, the temperature field calculation results include: internal cable core temperature data of the target cable and corresponding external environmental parameters, load parameters and time series.
[0082] S3. Construct a cable core temperature time series data set based on the parameter information and the temperature field calculation results of the cable finite element model.
[0083] As an optional implementation, the cross-sectional area, thickness, thermal conductivity of the target cable core, cable core shielding layer, insulation layer, metal shielding layer, and outer sheath layer, soil thermal resistance, soil temperature, external temperature, and convection heat transfer coefficient; cable core conductivity and relative magnetic permeability; and cable load parameters are used as data features, and the cable core temperature is used as a prediction variable to construct a cable core temperature time series data set.
[0084] S4. Using the Laplace fractional algorithm, the input features in the cable core temperature time series data set are evaluated for feature importance, and the N input features with the largest contribution are selected as the final input features of the model; N is an integer greater than 1.
[0085] In some embodiments, step S4 specifically includes the following steps:
[0086] S41. Assume that the cable core temperature time series data set has m samples, r features, and L r is the Laplace score of the rth feature, f ri is the rth feature of the i-th sample.
[0087] S42. Construct a nearest neighbor graph G with m nodes, where the i-th node corresponds to sample x i ; If the sample x i and sample x j Very close in the sample space, that is, sample x i Located in sample x j Among the k nearest neighbors, or sample x j Located in sample x i If , then there is an edge between nodes i and j in the nearest neighbor graph G; otherwise, nodes i and j are not connected.
[0088] S43, construct a weight matrix S, where The parameter t is a hyperparameter, ||·|| represents the Euclidean distance, and S ij Represents each element in the weight matrix S, and the weight matrix S can represent the topological structure of the data space.
[0089] S44. Calculate the Laplace matrix L=DS according to the weight matrix S, where D is the diagonal matrix D=diag(S) generated by the weight matrix S, and obtain the Laplace score of the rth feature:
[0090]
[0091] In the formula, f r =[f r1 , f r2 , …, f rm ] T ,
[0092] S45. Obtain N features according to the obtained Laplace scores as final input features of the model.
[0093] The smaller the Laplace score is, the more local information this feature can retain and the better the discriminability is, which means that this feature is more important.
[0094] S5. Based on the feature-selected cable core temperature time series data set, the long short-term memory network (LSTM) optimized by genetic algorithm is trained to obtain a cable temperature field prediction model for predicting the cable core temperature.
[0095] In some embodiments, after feature selection processing based on the Laplace fractional algorithm, the cable temperature time series data set is divided into 80% training set and 20% test set.
[0096] Specifically, the LSTM network consists of a forget gate, an input gate and an output gate.
[0097] The forget gate determines which information needs to be removed from the cell state. Its output is a number between 0 and 1, indicating the probability that some information is forgotten, calculated as:
[0098] f t =σ(W xf *x t +W lf *h t-1 +b f )
[0099] In the formula, f t Represents the output value of the forget gate at the current moment, x t Indicates the input at the current moment, h t-1 represents the hidden state at the previous moment, W xf and W hf Represents the input vector x t and the hidden state h t-1 and the weight matrix of the forget gate, b f represents the bias vector of the forget gate, and σ represents the Sigmoid function.
[0100] The input gate determines what new information needs to be added to the cell state. The calculation formula is:
[0101]
[0102] In the formula, i t represents the input gate output value at the current moment, c t-1 Indicates the cell state at the previous moment, Represents a new candidate cell state, W xi and W hi Represents the input vector x t and the hidden state h t-1 and the weight matrix of the input gate, b i represents the bias vector of the input gate, and tanh represents the hyperbolic tangent function.
[0103] The output gate determines which part of the cell state will be output to the hidden state. The calculation formula is:
[0104] o t =σ(W xo *x t +W ho *h t-1 +b o )
[0105] h t =o t *tanh(c t )
[0106] In the formula, o t Indicates the output value of the output gate at the current moment, o t represents the hidden state at the current moment, W xo and W ho Represents the input vector x t and the hidden state h t-1 and the weight matrix of the output gate, b o Represents the bias vector of the input gate.
[0107] As an optional implementation, the long short-term memory network is optimized in the following manner:
[0108] Genetic algorithms are used to optimize the number of layers of the long short-term memory network, the number of neurons in each layer, the number of hidden layers, and the learning rate. When the fitness function requirements are met, the optimization results, i.e., the optimal parameters of the model, are obtained.
[0109] The expression of the fitness function is:
[0110]
[0111] Where n represents the number of cable core temperature time series sample data in the test set, y i represents the real settlement data in the test set, represents the cable core temperature data predicted by the LSTM network, A represents a chromosome in the population, and R represents the preset hyperparameter value space.
[0112] The above method is supplemented and explained below with reference to the accompanying drawings and specific embodiments.
[0113] This embodiment provides a cable core temperature prediction method based on finite element method and Laplace fraction, which specifically includes the following steps:
[0114] Step 1: Construct a finite element model of the cable temperature field and perform finite element temperature field analysis and calculation on the target cable model under different operating conditions to obtain the time series data of the cable core temperature inside the target cable.
[0115] Step 1 specifically includes the following steps:
[0116] Step 1.1: Obtain the actual structural parameters, thermal parameters, electromagnetic parameters, laying conditions and boundary conditions of the target cable to establish a finite element model of the target cable.
[0117] Step 1.2: Based on the target cable finite element model, simulate the operating status under different working conditions, and perform finite element temperature field analysis and calculation to obtain the temperature data and time series of the target cable core.
[0118] Step 2: Based on the constructed cable temperature field finite element model and the acquired internal cable core temperature time series data of the target cable, a cable core temperature time series data set is established.
[0119] See also Figure 2 , Figure 3 and Figure 4 Specifically, according to the cable finite element model described in step 1, the cross-sectional area, thickness, thermal conductivity, soil thermal resistance, soil temperature, ambient temperature and convection heat transfer coefficient of the target cable's cable core, cable core shielding layer, insulation layer, metal shielding layer, and outer sheath layer; cable core conductivity and relative magnetic permeability; and cable load parameters are used as data features. The cable core temperature is used as a prediction variable to construct a cable core temperature time series data set.
[0120] Step 3: Use Laplace scores to perform feature selection on the data set to obtain the cable core temperature time series data set for the final input model, and divide it into training sets and test sets according to the proportion for the input of subsequent prediction models.
[0121] Step 3 specifically includes the following steps:
[0122] Step 3.1: Use the Laplace fractional algorithm to estimate the importance of the features of the cable core temperature time series data set described in step 2.
[0123] Step 3.2: Based on the Laplace scores of each feature, select the N features with the lowest scores as the final input features of the model.
[0124] The specific calculation process is as follows: Assuming that the cable core temperature time series data set has m samples, r features, L r is the Laplace score of the rth feature, f ri is the rth feature of the i-th sample. Construct a nearest neighbor graph G with m nodes, where the i-th node corresponds to sample x i If x i and x j is very close in the sample space, that is, x i Located at x j Among the k nearest neighbors of x j Located at x i Among the k nearest neighbors, there is an edge between nodes i and j; otherwise, nodes i and j are not connected. Construct a weight matrix S, where The parameter t is a hyperparameter, ||·|| represents the Euclidean distance, and S ij Represents each element in the weight matrix S, which can represent the topological structure of the data space. According to the weight matrix S, the Laplace matrix L = DS is calculated, where D is the diagonal matrix D = diag (S) generated by S. The Laplace score of the rth feature can be obtained as:
[0125]
[0126] Among them, f r =[f r1 , f r2 , …, f rm ] T ,
[0127] Step 3.3: After feature selection based on the Laplace fractional algorithm, the cable temperature time series dataset is divided into 80% training set and 20% test set.
[0128] Step 4: Build and initialize the LSTM network, and use the genetic algorithm to optimize the parameters of the LSTM network to meet the requirements of the fitness function and determine the optimal parameters of the model.
[0129] See also Figure 5 , the LSTM network consists of forget gate, input gate and output gate units. The forget gate determines which information needs to be removed from the cell state. Its output is a number between 0 and 1, indicating the probability of some information being forgotten, calculated as:
[0130] f t =σ(W xf *x t +W lf *h t-1 +b f )
[0131] In the formula, f t Represents the output value of the forget gate at the current moment, x t Indicates the input at the current moment, h t-1 represents the hidden state at the previous moment, W xf and W hf Represents the input vector x t and the hidden state h t-1 and the weight matrix of the forget gate, b f represents the bias vector of the forget gate, and σ represents the Sigmoid function.
[0132] The input gate determines what new information needs to be added to the cell state. The calculation formula is:
[0133]
[0134] In the formula, i t represents the input gate output value at the current moment, c t-1 Indicates the cell state at the previous moment, Represents a new candidate cell state, Wxi and W hi Represents the input vector x t and the hidden state h t-1 and the weight matrix of the input gate, b i represents the bias vector of the input gate, and tanh represents the hyperbolic tangent function.
[0135] The output gate determines which part of the cell state will be output to the hidden state. The calculation formula is:
[0136] o t =σ(W xo *x t +W ho *h t-1 +b o )
[0137] h t =o t *tanh(c t )
[0138] In the formula, o t Indicates the output value of the output gate at the current moment, o t represents the hidden state at the current moment, W xo and W ho Represents the input vector x t and the hidden state h t-1 and the weight matrix of the output gate, b o Represents the bias vector of the input gate.
[0139] The model optimization process is as follows:
[0140] In the genetic algorithm, the definition of chromosome representation is based on the set hyperparameter value space R. Each parameter is binary encoded, and the binary codes of these parameters are combined to form a complete chromosome individual. Subsequently, the population is initialized, which is the set of all generated chromosomes. Each individual in the population represents a potential optimal parameter combination. The population is initialized according to the range of hyperparameter values to accelerate its convergence to the global optimal solution. Finally, the fitness function used to evaluate the individual is determined. This function is used to evaluate the prediction accuracy of the model. The specific formula is:
[0141]
[0142] Where n represents the number of cable core temperature time series sample data in the test set, y i represents the real settlement data in the test set, represents the cable core temperature data predicted by the LSTM network, A represents a chromosome in the population, and R represents the preset hyperparameter value space.
[0143] Select excellent chromosomes from the population for reproduction, and use the tournament selection strategy for selection. Specifically, randomly select several individuals from the population to form a group, compare these individuals, and select the individual with the highest fitness as the parent individual for reproduction. This process is repeated until m individuals are selected for subsequent genetic operations. Set the crossover probability to 0.7, and pair the m individuals selected in the step without duplication. During the pairing process, generate a random number in the interval [0, 1]. If the random number is less than the crossover probability, a crossover operation is performed, otherwise the two chromosomes remain the same. For the chromosomes that need to be crossover, randomly determine the crossover point position, and exchange the rear part of the paired chromosome at the crossover point. Set the mutation probability to 0.05, and generate a random number in the interval [0, 1] for each of the new m individuals obtained in the step. If the generated random number is less than the mutation probability, a mutation operation is performed, otherwise the individual remains the same. The chromosome selected for the mutation operation randomly selects a gene bit in its binary code to invert, that is, 0 is changed to 1, and 1 is changed to 0. The mutated chromosome will be used as an individual in the offspring. Repeat the above crossover and mutation steps until the preset number of genetic iterations is reached or the fitness function reaches the target value. Finally, the chromosome with the best fitness is selected from the population. This chromosome represents a set of optimal hyperparameter values. The best hyperparameters are used to train the LSTM network, and the best weights obtained from the training are retained.
[0144] Step 5: Input the final input data set into the model described in Step 4 to complete the cable core temperature prediction.
[0145] Specifically, the first i-1 data points of the cable core temperature time series data sequence are used to predict the i-th data point, and the predicted value is added to the cable core temperature time series data sequence. At this time, the time series data sequence has a total of i data points, and then the first data point is removed. Next, the last i-1 data points of the updated time series data sequence are used to predict the i+1 data point. By repeating the above steps, the prediction of the cable core time series temperature is achieved.
[0146] In summary, the embodiment of the present invention provides a cable core temperature prediction method based on the finite element method and Laplace fraction, including: using the cable's structural parameters, electromagnetic parameters, thermal parameters and boundary conditions to perform finite element analysis and calculation, and obtaining the theoretical influencing factors and cable core temperature time series data of the cable under different operating conditions; based on the Laplace fraction algorithm, the theoretical influencing factors are evaluated for feature importance to form a multidimensional feature set of the cable core temperature; using a genetic algorithm to optimize the parameters of the long short-term memory (LSTM) network to construct a cable core temperature field prediction model. The present invention adopts a feature selection algorithm based on an interpretable graph operator to realize the cable core temperature prediction based on interpretable machine learning, which not only explains the nonlinear relationship between the cable core temperature and the prediction variable, but also avoids the singularity problem that occurs when calculating the projection matrix, effectively enhancing the utilization of data, and improving the accuracy of the model, providing theoretical support and new ideas for constructing interpretable cable core temperature prediction.
[0147] This embodiment also provides a cable core temperature prediction device, comprising:
[0148] at least one processor;
[0149] at least one memory for storing at least one program;
[0150] When the at least one program is executed by the at least one processor, the at least one processor implements the following Figure 1 The method shown.
[0151] It is understood that the memory may include a random access memory (RAM) or a read-only memory (ROM). Optionally, the memory includes a non-transitory computer-readable storage medium. The memory may be used to store instructions, programs, codes, code sets, or instruction sets. The memory may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function, instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area may store data created according to the use of the server, etc.
[0152] The processor may include one or more processing cores. The processor uses various interfaces and lines to connect the various parts of the entire server, and executes various functions of the server and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory, and calling data stored in the memory. Optionally, the processor can be implemented in at least one hardware form of digital signal processing (DSP), field-programmable gate array (FPGA), and programmable logic array (PLA). The processor can integrate one or a combination of a central processing unit (CPU) and a modem. Among them, the CPU mainly processes the operating system and application programs; the modem is used to process wireless communications. It can be understood that the above-mentioned modem may not be integrated into the processor, but implemented separately through a chip.
[0153] Since the device is a device corresponding to a cable core temperature prediction method in an embodiment of the present invention, and the principle of solving the problem by the device is similar to that of the method, the implementation of the device can refer to the implementation process of the above method embodiment, and the repeated parts will not be repeated.
[0154] The embodiment of the present invention further provides a computer-readable storage medium, wherein the storage medium stores at least one instruction, at least one program, code set or instruction set, and the at least one instruction, the at least one program, the code set or instruction set is loaded and executed by a processor to implement the following Figure 1 A cable core temperature prediction method is shown in FIG.
[0155] Those skilled in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable rewritable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.
[0156] Since the storage medium is a storage medium corresponding to a cable core temperature prediction method in an embodiment of the present invention, and the principle of solving the problem by the storage medium is similar to that of the method, the implementation of the storage medium can refer to the implementation process of the above-mentioned method embodiment, and the repeated parts will not be repeated.
[0157] In some possible implementations, various aspects of the method of the embodiment of the present invention can also be implemented in the form of a program product, which includes a program code. When the program product is run on a computer device, the program code is used to enable the computer device to execute the steps of a cable core temperature prediction method according to various exemplary embodiments of the present application described above in this specification. Among them, the executable computer program code or "code" for executing each embodiment can be written in a high-level programming language such as C, C++, C#, Smalltalk, Java, JavaScript, Visual Basic, structured query language (e.g., Transact-SQL), Perl, or in various other programming languages.
[0158] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.
[0159] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.
[0160] The above embodiments are only for illustrating the technical concept and features of the present invention, and their purpose is to enable ordinary technicians in the field to understand the content of the present invention and implement it accordingly, and they cannot be used to limit the protection scope of the present invention. Any equivalent changes or modifications made based on the essence of the content of the present invention should be included in the protection scope of the present invention.
Claims
1. A method for predicting cable core temperature. It is characterized in that The following steps are involved: Obtain parameter information of the target cable and establish a finite element model of the target cable; Under different operating conditions, the finite element temperature field analysis and calculation are performed on the target cable finite element model to obtain the temperature field calculation results; According to the parameter information and the temperature field calculation results of the cable finite element model, a cable core temperature time series data set is constructed; The Laplace fractional algorithm is used to evaluate the feature importance of the input features in the cable core temperature time series data set, and multiple input features with the largest contribution are selected as the final input features of the model; Based on the feature-selected cable core temperature time series data set, the long short-term memory network optimized by genetic algorithm is trained to obtain the cable temperature field prediction model, which is used to predict the cable core temperature.
2. A cable core temperature prediction method according to claim 1, It is characterized in that The parameter information of the target cable includes: structural parameters, thermal parameters, electromagnetic parameters, laying conditions and boundary conditions of the target cable; The temperature field calculation results include: the internal cable core temperature data of the target cable and the corresponding external environment parameters, load parameters and time series; The structural parameters, thermal parameters, electromagnetic parameters, laying conditions, external environmental parameters, load parameters and time series of the target cable are taken as the input characteristics of the model, and the cable core temperature is taken as the output variable of the model.
3. A cable core temperature prediction method according to claim 1 or 2, It is characterized in that The parameter information of the target cable includes: The cross-sectional area and thickness of the cable core, cable core shielding layer, insulation layer, metal shielding layer and outer sheath layer of the target cable; and, Thermal conductivity of the cable core, cable core shielding layer, insulation layer, metal shielding layer and outer sheath layer of the target cable; and, soil thermal resistance, soil temperature, ambient temperature and convection heat transfer coefficient; and, Cable core conductivity and relative magnetic permeability; and, Load parameters of the cable.
4. A cable core temperature prediction method according to claim 1, It is characterized in that The Laplace fractional algorithm is combined to evaluate the feature importance of the input features in the cable core temperature time series data set, and multiple input features with the largest contribution are selected as the final input features of the model, including: Assume that the cable core temperature time series dataset has m samples, r features, and L r is the Laplace score of the rth feature, f ri is the rth feature of the i-th sample; Construct a nearest neighbor graph G with m nodes, where the i-th node corresponds to sample x i ; If the sample x i and sample x j Very close in the sample space, that is, sample x i Located in sample x j Among the k nearest neighbors, or sample x j Located in sample x i Among the k nearest neighbors, there is an edge between nodes i and j in the nearest neighbor graph G; otherwise, nodes i and j are not connected; Construct the weight matrix S, where The parameter t is a hyperparameter, ∥·∥ represents the Euclidean distance, and S ij Represents each element in the weight matrix S; The Laplace matrix L=DS is calculated based on the weight matrix S, where D is a diagonal matrix generated by the weight matrix S, and the Laplace score of the rth feature is obtained as follows: In the formula, f r =[f r1 ,f r2 ,…,f rm ] T , Multiple features are obtained according to the obtained Laplace scores as the final input features of the model.
5. A cable core temperature prediction method according to claim 1, It is characterized in that The long short-term memory network is composed of a forget gate, an input gate and an output gate; The forget gate is used to determine whether information should be removed from the cell state. Its output is a number between 0 and 1, indicating the probability of information being forgotten. The calculation formula is: f t =σ(W xf *x t +W lf *h t-1 +b f ) In the formula, f t Represents the output value of the forget gate at the current moment, x t Indicates the input at the current moment, h t-1 represents the hidden state at the previous moment, W xf and W hf Represents the input vector x t and the hidden state h t-1 and the weight matrix of the forget gate, b f represents the bias vector of the forget gate, σ represents the Sigmoid function; The input gate is used to determine whether new information needs to be added to the cell state. The calculation formula is: In the formula, i t represents the input gate output value at the current moment, c t-1 Indicates the cell state at the previous moment, Represents a new candidate cell state, W xi and W hi Represents the input vector x t and the hidden state h t-1 and the weight matrix of the input gate, b i represents the bias vector of the input gate, and tanh represents the hyperbolic tangent function; The output gate is used to determine which part of the cell state will be output to the hidden state, and the calculation formula is: o t =σ(W xo *x t +W ho *h t-1 +b o ) h t =o t *tanh(c t ) In the formula, o t Indicates the output value of the output gate at the current moment, o t represents the hidden state at the current moment, W xo and W ho Represents the input vector x t and the hidden state h t-1 and the weight matrix of the output gate, b o Represents the bias vector of the input gate.
6. A cable core temperature prediction method according to claim 1, It is characterized in that The long short-term memory network is optimized in the following way: Genetic algorithms are used to optimize the number of layers of the long short-term memory network, the number of neurons in each layer, the number of hidden layers, and the learning rate. When the fitness function requirements are met, the optimization results, i.e., the optimal parameters of the model, are obtained. The expression of the fitness function is: Where n represents the number of cable core temperature time series sample data in the test set, y i represents the real settlement data in the test set, represents the cable core temperature data predicted by the LSTM network, A represents a chromosome in the population, and R represents the preset hyperparameter value space.
7. A cable core temperature prediction method according to claim 1 or 6, It is characterized in that The working mode of the genetic algorithm optimization is: Initialize the population; Repeatedly randomly select a number of individuals from the population to form a group, compare these individuals, and select the individual with the highest fitness as the parent individual for reproduction until m individuals are selected for subsequent genetic operations; Crossover step: pair the selected m individuals in pairs without duplication, and generate a random number in the interval [0,1] during the pairing process. If the random number is less than the preset crossover probability, a crossover operation is performed; Otherwise, the two chromosomes remain intact; Mutation step: For each of the m individuals that have undergone the crossover step, generate a random number in the interval [0,1]. If the generated random number is less than the preset mutation probability, a mutation operation is performed; otherwise, the individual remains the same; the mutated chromosome will be used as an individual in the offspring; Repeat the crossover step and the mutation step until the preset number of genetic iterations is reached or the fitness function reaches the target value; The chromosome with the best fitness is selected from the population. The chromosome represents a set of optimal hyperparameter values. The LSTM network is trained with the optimal hyperparameter values, and the best weights obtained by training are retained.
8. A cable core temperature prediction method according to claim 1, It is characterized in that The method of obtaining a cable temperature field prediction model for predicting the cable core temperature includes: Obtain the first i-1 data points of the cable core temperature time series data sequence, input them into the cable temperature field prediction model, predict and obtain the i-th data point, and add the obtained data point to the cable core temperature time series data sequence; Remove the first data point in the cable core temperature time series data sequence, input the i-1 data point of the updated cable core temperature time series data sequence into the cable temperature field prediction model, and predict the i+1th data point; The newly obtained data point is added to the cable core temperature time series data sequence, and the executed step of removing the first data point in the cable core temperature time series data sequence is returned.
9. A cable core temperature prediction device, It is characterized in that include: at least one processor; at least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements the method according to any one of claims 1 to 8.
10. A computer-readable storage medium having a program executable by a processor stored therein, It is characterized in that The processor-executable program is used to perform the method according to any one of claims 1 to 8 when executed by the processor.
Citation Information
Cited By
Intelligent monitoring system and method based on artificial intelligence
CN120538578A
Photovoltaic cable extrusion control method
CN121246206A
