Cable temperature sensing method, medium and electronic equipment

Abstract

The invention discloses a cable temperature sensing method, a medium and electronic equipment. The method comprises the following steps: calculating a cable temperature data set by adopting a finite element calculation method; calculating the importance score of each characteristic variable influencing the cable temperature in a random forest by adopting a random forest method, screening out at least one importance characteristic variable as an input quantity according to the importance score, and solving a pre-established support vector regression model by taking the cable temperature data set as an output quantity, obtaining an optimal weight vector and an optimal temperature deviation value in the support vector regression machine model; constructing a cable temperature virtual sensing model based on the optimal weight vector and the optimal temperature deviation value; and monitoring the running state of the cable in real time, and obtaining parameter values of the plurality of importance characteristic variables so as to realize real-time sensing of the cable temperature according to the parameter values of the plurality of importance characteristic variables and the cable temperature virtual sensing model. The method is realized based on a random forest method and a support vector regression machine model, and the cable temperature can be accurately sensed in real time.

Description

Technical Field

[0001] The present invention relates to the technical field of cable temperature measurement, and in particular to a cable temperature sensing method, medium and electronic equipment. Background Art

[0002] Digital twin technology collects and analyzes data from physical entities, and establishes a digital twin model in the information technology platform that has the same physical operating characteristics and principles as the entity. The digital twin model can be corrected based on the data collected in real time to achieve synchronization with the actual physical entity. The digital twin model is used to simulate and predict specific working conditions, fault warnings, etc., to achieve optimal decision-making and control of the physical entity. This technology can truly reflect the characteristics and working principles of the physical entity. With the help of digital twin technology, digital models and physical entities can be synchronized in real time and have the ability to interact in virtual reality.

[0003] With the widespread application of power cables in the field of power transmission and distribution, and the power sector's increasing attention to the reliability, safety and stability of power cable operation, how to more accurately and quickly analyze the temperature distribution of power cables and how to evaluate the operating status of power cables have become research hotspots.

[0004] The calculation method of cable temperature generally adopts numerical method. Numerical method calculates the whole temperature field under given cable electromagnetic parameters and environmental parameters, including finite element calculation method, which has the advantages of simulating complex models and high calculation accuracy. However, under the same cable laying conditions, the time of calculating cable temperature by finite element method increases. When constructing the cable, the multi-physics field coupling model is applied to the digital twin cable framework. The model-driven theory represented by finite element calculation method cannot meet the real-time requirements of digital twin applications, so it is impossible to accurately and virtually perceive the cable temperature in real time. Summary of the invention

[0005] The present invention aims to solve at least one of the technical problems in the related art to a certain extent. To this end, the first object of the present invention is to provide a cable temperature sensing method, which is based on a random forest method and a support vector regression model and can accurately and real-time sense the cable temperature.

[0006] A second object of the present invention is to provide a computer-readable storage medium.

[0007] A third object of the present invention is to provide an electronic device.

[0008] To achieve the above object, the present invention is implemented through the following technical solutions:

[0009] A cable temperature sensing method, comprising:

[0010] The cable temperature data set is obtained by using the finite element calculation method;

[0011] The importance scores of the characteristic variables affecting the cable temperature in the random forest are calculated by using the random forest method, at least one important characteristic variable is selected as an input according to the importance score, and the pre-established support vector regression model is solved with the cable temperature data set as the output to obtain the optimal weight vector and the optimal temperature deviation value in the support vector regression model;

[0012] Constructing a cable temperature virtual perception model based on the optimal weight vector and the optimal temperature deviation value;

[0013] The cable operation status is monitored in real time, and parameter values ​​of a plurality of the important characteristic variables are obtained, so as to realize real-time perception of the cable temperature according to the parameter values ​​of the plurality of important characteristic variables and the cable temperature virtual perception model.

[0014] Preferably, the cable temperature data set is calculated using a finite element calculation method, including:

[0015] Build a relationship model between cable temperature and total cable loss, and set cable boundary conditions, where the cable boundary conditions are cable surface heat dissipation conditions;

[0016] Constructing a finite element calculation model based on the relationship model between the cable temperature and the total cable loss and the cable boundary conditions;

[0017] Ambient temperature parameters, load parameters and initial cable temperature parameters are input into the finite element calculation model to obtain the cable temperature data set.

[0018] Preferably, the relationship model between the cable temperature and the total cable loss is expressed as follows:

[0019]

[0020] Q h =W e +W d +W h

[0021] in, represents the divergence operator, k represents the thermal conductivity, represents the gradient operator, T represents the cable temperature, Q h Represents the total cable loss, W e Indicates the cable core loss, W d Indicates the dielectric loss, W h Indicates the cable circulation loss.

[0022] Preferably, the cable boundary condition is expressed as follows:

[0023]

[0024] Where v is the cable length in the heat flow direction, Γ is the integration boundary, α is the cable surface convection heat transfer coefficient, T f is the outside air temperature.

[0025] Preferably, the characteristic variables include load current, cable core conductivity, ambient temperature, cable surface convection heat transfer coefficient and measurement point position in the cable.

[0026] Preferably, the random forest method is used to calculate the importance scores of the characteristic variables affecting the cable temperature in the random forest, including:

[0027] Determine all nodes in a decision tree that include the feature variable;

[0028] Calculate the Gini index of each node, and calculate the importance score of the feature variable in each node based on the Gini index of each node;

[0029] Calculating the importance score of the feature variable in the decision tree based on the importance score of the feature variable in each node;

[0030] The importance scores of the feature variables in each decision tree are determined to determine the importance scores of the feature variables in a random forest including each of the decision trees.

[0031] Preferably, the support vector regression model is expressed as follows:

[0032]

[0033] Among them, min is the minimum function, w is the weight vector, b is the temperature deviation value, ξ i , i * are the upper and lower tolerance errors of the i-th sample, C is the penalty coefficient, s is the number of samples, and y i is the target temperature value of the i-th sample, φ(x i ) is the kernel function, x i is the i-th sample, ε is the insensitive loss function threshold, where each sample is a feature vector composed of multiple important feature variables.

[0034] Preferably, the cable temperature virtual perception model is expressed as follows:

[0035] f(x)= <w * ,φ(x i )>+b *

[0036] Where f(x) represents the real-time sensed value of cable temperature, w * 、b * Represent the optimal weight vector and optimal temperature deviation value respectively.

[0037] To achieve the above-mentioned object, the second aspect of the present invention provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the above-mentioned cable temperature sensing method is implemented.

[0038] To achieve the above objectives, the third aspect of the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the cable temperature sensing method described above is implemented.

[0039] The present invention has at least the following technical effects:

[0040] The present invention provides a cable temperature perception method based on a random forest method and a support vector regression model. The method uses a finite element calculation method to calculate a cable temperature data set, then uses a random forest method to calculate the importance score of each characteristic variable affecting the cable temperature in the random forest, and selects at least one important characteristic variable as an input quantity, and then uses the cable temperature data set as an output quantity to solve a pre-established support vector regression model, obtains the parameters in the support vector regression model, namely, an optimal weight vector and an optimal temperature deviation value, and then builds a cable temperature virtual perception model based on the optimal weight vector and the optimal temperature deviation value, and then only needs to monitor the cable operation state in real time, obtain the parameter values ​​of multiple important characteristic variables, and then the cable temperature real-time virtual perception can be realized according to the parameter values ​​of multiple important characteristic variables and the cable temperature virtual perception model. Therefore, the present invention does not need to embed a sensor to monitor the cable temperature, and can perform real-time virtual perception of the cable temperature based on the cable temperature virtual perception model constructed by the random forest method and the support vector regression model, which lays a foundation for further accurately evaluating the operation state of the cable, and has important theoretical and application value for improving the cable safety level and the accuracy of operation state evaluation.

[0041] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 The figure is a flow chart of a cable temperature sensing method according to an embodiment of the present invention.

[0043] Figure 2 The figure is a specific flow chart of the cable temperature sensing method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0044] The present embodiment is described in detail below, and examples of the embodiment 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 intended to be used to explain the present invention, and should not be construed as limiting the present invention.

[0045] The cable temperature sensing method, medium and electronic device of this embodiment are described below with reference to the accompanying drawings.

[0046] Figure 1 FIG. 1 is a flow chart of a cable temperature sensing method according to an embodiment of the present invention. Figure 1 As shown, the method includes:

[0047] Step S1: Calculate the cable temperature data set using the finite element calculation method.

[0048] In this embodiment, a finite element calculation method is used to calculate the cable temperature data set, including: building a relationship model between cable temperature and total cable loss, setting cable boundary conditions, and the cable boundary conditions are the heat dissipation conditions on the cable surface; constructing a finite element calculation model based on the relationship model between cable temperature and total cable loss and the cable boundary conditions; inputting ambient temperature parameters, load parameters and initial cable temperature parameters into the finite element calculation model to obtain a cable temperature data set.

[0049] Specifically, the cable temperature can be obtained by using a finite element calculation method. In this embodiment, the finite element calculation method comprises the following steps: first, a relationship model between the cable temperature and the total cable loss is constructed in COMSOL Multiphysics (a simulation software). The relationship model between the cable temperature and the total cable loss is expressed as follows:

[0050]

[0051] Q h =W e +W d +W h (2)

[0052] in, represents the divergence operator, k represents the thermal conductivity, represents the gradient operator, T represents the cable temperature, Q h Represents the total cable loss, W e Indicates the cable core loss, W d Indicates the dielectric loss, W h Indicates the cable circulation loss.

[0053] Furthermore, a cable boundary condition is set, which is a cable surface heat dissipation condition. The cable boundary condition is expressed as follows:

[0054]

[0055] Where v is the cable length in the heat flow direction, Γ is the integration boundary, α is the cable surface convection heat transfer coefficient, T f is the outside air temperature.

[0056] Further, such as Figure 2 As shown, the cable structure parameters and material thermal parameters are obtained, and then a finite element calculation model is constructed in the software based on the relationship model between cable temperature and total cable loss and the set boundary conditions. Then, the ambient temperature parameters, load parameters and initial cable temperature parameters are input into the finite element calculation model to obtain the cable temperature data set.

[0057] Step S2: The random forest method is used to calculate the importance scores of the characteristic variables affecting the cable temperature in the random forest, and at least one important characteristic variable is selected as the input according to the importance score. The cable temperature data set is used as the output to solve the pre-established support vector regression model to obtain the optimal weight vector and the optimal temperature deviation value in the support vector regression model.

[0058] In this embodiment, the characteristic variables include load current, cable core conductivity, ambient temperature, cable surface convection heat transfer coefficient, and measurement point position in the cable.

[0059] Among them, the random forest method, namely the RF algorithm, is used to calculate the importance scores of the characteristic variables affecting the cable temperature in the random forest, including: determining all nodes in a decision tree including the characteristic variables; calculating the Gini (an indicator for measuring the purity of a data set) index of each node, and calculating the importance scores of the characteristic variables in each node based on the Gini index of each node; calculating the importance scores of the characteristic variables in the decision tree based on the importance scores of the characteristic variables in each node; determining the importance scores of the characteristic variables in each decision tree, so as to determine the importance scores of the characteristic variables in the random forest including each decision tree.

[0060] Assuming that there are feature variables X1, X2, ..., Xj, where Xj represents the jth feature variable, the steps for calculating the importance score of Xj in the random forest are as follows:

[0061] In a decision tree, the Gini index of node m is expressed as follows:

[0062]

[0063] Among them, G mis the Gini index of node m; K is the number of categories of all feature variables at node m; P mr is the probability estimate of the feature variable belonging to category r at node m.

[0064] The importance score of feature variable Xj in node m, that is, the calculation formula of the change of Gini index before and after node m is split, is:

[0065]

[0066] in, is the importance score of feature variable Xj in node m, G ml and G mr Respectively represent the Gini index of the left node and the right node split by node m.

[0067] The importance score calculation formula of feature variable Xj in decision tree u is as follows:

[0068]

[0069] in, is the importance score of feature variable Xj in decision tree u; M is the number of nodes in decision tree u that contain feature variable Xj.

[0070] The calculation formula for the importance score of feature variable Xj in random forest is as follows:

[0071]

[0072] in, is the importance score of feature variable Xj in random forest; U is the number of decision trees in random forest.

[0073] Furthermore, multiple important characteristic variables are selected according to the importance scores as input, and the corresponding cable temperature is taken from the cable temperature data set as the output to solve the following pre-established support vector regression model. For example, the multiple important characteristic variables selected are X1, X2, and X4, and a set of X1, X2, and X4 values ​​are taken to form the i-th sample x i And as the input, take the cable temperature value obtained by the combined influence of the group of X1, X2, and X4 values ​​as the i-th sample x i The target temperature value y i When X1, X2, and X4 take another set of values, it obviously corresponds to another sample, that is, another input quantity, and another target temperature value in the cable temperature data set.

[0074] The SVR model, i.e., the support vector regression model, is established in advance. The support vector regression model is expressed as follows:

[0075]

[0076] Among them, min is the minimum function, w is the weight vector, b is the temperature deviation value, ξ i , i * are the upper and lower tolerance errors of the i-th sample, C is the penalty coefficient, s is the number of samples, and y i is the target temperature value of the i-th sample, φ(x i ) is the kernel function, x i is the i-th sample, ε is the insensitive loss function threshold, where each sample is a feature vector composed of multiple important feature variables.

[0077] The kernel function is expressed as follows:

[0078]

[0079] Among them, γ represents the bandwidth parameter of the kernel function, x σ represents the σth sample, that is, another sample, ||x i -x σ || represents the Euclidean distance between two samples.

[0080] Furthermore, the input and output are input and the support vector regression algorithm is used to solve the above support vector regression model to obtain a set of optimal solutions w * 、b * .w * is the optimal weight vector obtained by solving, b * is the optimal temperature deviation value obtained by solving.

[0081] Step S3: constructing a cable temperature virtual perception model based on the optimal weight vector and the optimal temperature deviation value.

[0082] The cable temperature virtual perception model is expressed as follows:

[0083] f(x)= <w * ,φ(x i )>+b * (10)

[0084] Where f(x) represents the real-time sensed value of cable temperature, w * 、b * Represent the optimal weight vector and optimal temperature deviation value respectively.

[0085] Step S4: monitor the cable operation status in real time and obtain parameter values ​​of multiple important characteristic variables, so as to realize real-time cable temperature perception according to the parameter values ​​of multiple important characteristic variables and the cable temperature virtual perception model.

[0086] After finding multiple important characteristic variables through the random forest method, real-time monitoring is performed during the operation of the cable to obtain the values ​​of these important characteristic variables, which are then directly substituted into the cable temperature virtual perception model to directly realize real-time virtual perception of the cable temperature.

[0087] In summary, the present invention provides a cable temperature perception method based on random forest method and support vector regression model. The method adopts finite element calculation method to calculate the cable temperature data set, and then adopts random forest method to calculate the importance score of each characteristic variable affecting the cable temperature in the random forest, and selects at least one important characteristic variable as input, and then solves the pre-established support vector regression model with the cable temperature data set as output, obtains the parameters in the support vector regression model, namely the optimal weight vector and the optimal temperature deviation value, and then constructs a cable temperature virtual perception model based on the optimal weight vector and the optimal temperature deviation value, and then only needs to monitor the cable operation state in real time, obtain the parameter values ​​of multiple important characteristic variables, and then realize the real-time virtual perception of cable temperature according to the parameter values ​​of multiple important characteristic variables and the cable temperature virtual perception model. Therefore, the present invention does not need to embed sensors to monitor the cable temperature, and can perform real-time virtual perception of the cable temperature based on the cable temperature virtual perception model constructed by random forest method and support vector regression model, which lays the foundation for further accurate evaluation of the operation state of the cable, and has important theoretical and application value for improving the cable safety level and the accuracy of operation state evaluation.

[0088] Furthermore, the present invention also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the above-mentioned cable temperature sensing method can be implemented.

[0089] Furthermore, the present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the above-mentioned cable temperature sensing method can be implemented.

[0090] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0091] Although the content of the present invention has been described in detail through the above preferred embodiments, it should be appreciated that the above description should not be considered as a limitation of the present invention. After reading the above content, it will be apparent to those skilled in the art that various modifications and substitutions of the present invention will occur. Therefore, the protection scope of the present invention should be limited by the appended claims.

Claims

1. A cable temperature sensing method, characterized in that: include: The cable temperature data set is obtained by using the finite element calculation method; The importance scores of the characteristic variables affecting the cable temperature in the random forest are calculated by using the random forest method, at least one important characteristic variable is selected as an input according to the importance score, and the pre-established support vector regression model is solved with the cable temperature data set as the output to obtain the optimal weight vector and the optimal temperature deviation value in the support vector regression model; Constructing a cable temperature virtual perception model based on the optimal weight vector and the optimal temperature deviation value; The cable operation status is monitored in real time, and parameter values ​​of a plurality of the important characteristic variables are obtained, so as to realize real-time perception of the cable temperature according to the parameter values ​​of the plurality of important characteristic variables and the cable temperature virtual perception model.

2. The cable temperature sensing method according to claim 1, characterized in that: The cable temperature data set is calculated using the finite element method, including: Build a relationship model between cable temperature and total cable loss, and set cable boundary conditions, where the cable boundary conditions are cable surface heat dissipation conditions; Constructing a finite element calculation model based on the relationship model between the cable temperature and the total cable loss and the cable boundary conditions; Ambient temperature parameters, load parameters and initial cable temperature parameters are input into the finite element calculation model to obtain the cable temperature data set.

3. The cable temperature sensing method according to claim 2, characterized in that: The relationship model between the cable temperature and the total cable loss is expressed as follows: Q h =W e +W d +W h in, represents the divergence operator, k represents the thermal conductivity, represents the gradient operator, T represents the cable temperature, Q h Represents the total cable loss, W e Indicates the cable core loss, W d Indicates the dielectric loss, W h Indicates the cable circulation loss.

4. The cable temperature sensing method according to claim 3, characterized in that: The cable boundary conditions are expressed as follows: Where v is the cable length in the heat flow direction, Γ is the integration boundary, α is the cable surface convection heat transfer coefficient, T f is the outside air temperature.

5. The cable temperature sensing method according to claim 1, characterized in that: The characteristic variables include load current, cable core conductivity, ambient temperature, cable surface convection heat transfer coefficient and measurement point position in the cable.

6. The cable temperature sensing method according to claim 1, characterized in that: The random forest method is used to calculate the importance scores of the characteristic variables affecting the cable temperature in the random forest, including: Determine all nodes in a decision tree that include the feature variable; Calculate the Gini index of each node, and calculate the importance score of the feature variable in each node based on the Gini index of each node; Calculating the importance score of the feature variable in the decision tree based on the importance score of the feature variable in each node; The importance scores of the feature variables in each decision tree are determined to determine the importance scores of the feature variables in a random forest including each of the decision trees.

7. The cable temperature sensing method according to claim 1, characterized in that: The support vector regression model is expressed as follows: Among them, min is the minimum function, w is the weight vector, b is the temperature deviation value, ξ i , i * are the upper and lower tolerance errors of the i-th sample, C is the penalty coefficient, s is the number of samples, and y i is the target temperature value of the i-th sample, φ(x i ) is the kernel function, x i is the i-th sample, ε is the insensitive loss function threshold, where each sample is a feature vector composed of multiple important feature variables.

8. The cable temperature sensing method according to claim 7, characterized in that: The cable temperature virtual perception model is expressed as follows: f(x)=<w * ,φ(x i )>+b * Where f(x) represents the real-time sensed value of cable temperature, w * 、b * Represent the optimal weight vector and optimal temperature deviation value respectively.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the cable temperature sensing method according to any one of claims 1 to 8 is implemented.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the cable temperature sensing method according to any one of claims 1 to 8 is implemented.

Images