Conductive performance detection method and system for switch contact
By constructing a three-dimensional structure and a convolutional neural network to detect the conductivity of the contacts, the problem of inaccurate evaluation of the conductivity of the contacts in the prior art is solved, real-time and accurate evaluation of the conductivity of the contacts is achieved, and the reliability and safety of the equipment are improved.
Patent Information
- Application Number
- CN202510384748.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-28
- Publication Date
- 2025-08-01
AI Technical Summary
In the prior art, the conductive performance detection methods of contacts are mostly static tests, which cannot accurately reflect the changes in the conductive performance of the contacts in actual working conditions, affecting the reliability and safety of the equipment.
By obtaining the contact voltage drop, contact temperature and current matrix at multiple time points, a three-dimensional structure is constructed, and a convolutional neural network is used to detect conductivity and predict the time point at which the preset current value is reached, real-time and accurate evaluation of the contact conductivity performance is achieved.
Real-time and accurate evaluation of the conductive properties of the contact under different working conditions is achieved, and the reliability and safety of the equipment are improved.
Smart Images
Figure CN120405259A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of computer technology, and in particular, to a method and system for detecting the electrical conductivity of a contact for a switch. Background Art
[0002] Currently, in electrical equipment such as power switches, relays, and circuit breakers, the electrical conductivity of the contacts directly affects the working efficiency and safety of the equipment. As the usage time of the equipment increases, the contact surface may experience an increase in contact resistance due to corrosion, wear, etc., thus affecting the normal operation of the switch. However, the existing methods for detecting the electrical conductivity of contacts have certain limitations. Most of the existing detection methods are static tests and cannot fully reflect the change in the electrical conductivity of the contacts under actual working conditions. In order to improve the reliability and safety of the equipment, a new method and system for detecting electrical conductivity are needed, which can evaluate the electrical conductivity of the contacts in real time and accurately under different working conditions. Summary of the Invention
[0003] The purpose of the present invention is to provide a method and system for detecting the electrical conductivity of a contact for a switch to solve the above problems existing in the prior art.
[0004] In a first aspect, an embodiment of the present invention provides a method for detecting the electrical conductivity of a contact for a switch, including:
[0005] Obtaining the contact voltage drop and contact temperature at multiple time points; the contact voltage drop represents the voltage loss when the energized current passes through the contact for the switch; the contact temperature represents the temperature rise when the contacts for the switch are in contact;
[0006] Obtaining a current matrix at multiple time points; the current matrix includes a plurality of current differences;
[0007] Based on the multiple current matrices, constructing a three-dimensional structure, detecting the electrical conductivity, and obtaining a first electrical conductivity feature;
[0008] Based on the multiple current matrices, contact voltage drop, and contact temperature, detecting the electrical conductivity, and obtaining a second electrical conductivity feature; the second electrical conductivity feature represents the feature of the change in electrical conductivity corresponding to the current, voltage, and temperature at multiple time points;
[0009] Based on the first electrical conductivity feature and the second electrical conductivity feature, predicting the time point when a preset current value is reached, and obtaining the electrical conductivity time length.
[0010] Optionally, the step of based on the multiple current matrices, constructing a three-dimensional structure, detecting the electrical conductivity, and obtaining a first electrical conductivity feature includes:
[0011] Superimposing the multiple current matrices to obtain a three-dimensional current matrix;
[0012] Based on the three-dimensional current matrix, through a first convolutional network, detect the changes in all current differences at different times to obtain a first change feature;
[0013] The first convolutional network includes a plurality of first three-dimensional convolutional kernels;
[0014] Based on the three-dimensional current matrix, through a second convolutional network, detect the changes in the current differences between different energized currents and at different times to obtain a second change feature;
[0015] The second convolutional network includes a plurality of second three-dimensional convolutional kernels; the second convolutional network performs convolution along the diagonal;
[0016] Superimpose the first change feature and the second change feature to obtain a first conductivity feature;
[0017] Wherein, the first change feature, the second change feature, and the first conductivity feature are one-dimensional vectors.
[0018] Optionally, the detecting the conductivity based on a plurality of current matrices, contact voltage drops, and contact temperatures to obtain a second conductivity feature includes:
[0019] Sum the multiple values of the current matrix to obtain a total current difference;
[0020] Construct a conductivity relationship vector from the total current difference, contact temperature, and contact voltage drop; obtain a plurality of conductivity relationship vectors corresponding to multiple time points;
[0021] Input the plurality of conductivity relationship vectors into a temporal convolutional network in chronological order from early to late to obtain a second conductivity feature.
[0022] Optionally, the detecting the changes in the current differences between different energized currents and at different times based on the three-dimensional current matrix through a second convolutional network to obtain a second change feature includes:
[0023] Take the center point of the three-dimensional current matrix as the three-dimensional center point;
[0024] Based on the three-dimensional current matrix, obtain a first straight line, a second straight line, a third straight line, and a fourth straight line;
[0025] The second three-dimensional convolutional kernel is a 3*3*3 three-dimensional convolutional kernel;
[0026] Based on the first straight line, the second straight line, the third straight line, and the fourth straight line, perform convolution on the second three-dimensional convolutional kernel and the three-dimensional current matrix to obtain a second change feature.
[0027] Optionally, based on the first straight line, the second straight line, the third straight line, and the fourth straight line, convolving the second three-dimensional convolution kernel with the three-dimensional current matrix to obtain a second variation feature, including:
[0028] Obtain the center point of the second three-dimensional convolution kernel as the center point of the second three-dimensional convolution kernel;
[0029] Mark the positions where the three-dimensional current matrix coincides with the first straight line in the order of the first straight line to obtain a set of straight line positions;
[0030] Match the center point of the second three-dimensional convolution kernel with the values in the set of straight line positions, and perform convolution on the three-dimensional current matrix with a step size of 1 to obtain a second variation feature; the second variation feature is used to detect the features that change jointly with time and current.
[0031] Optionally, based on the three-dimensional current matrix, detecting the change of all current differences at different times through a first convolutional network to obtain a first variation feature, including:
[0032] The first three-dimensional convolution kernel is a three-dimensional convolution kernel of n*m*2; n represents the number of rows of the first three-dimensional convolution kernel; m represents the number of columns of the first three-dimensional convolution kernel; 2 corresponds to the number of layers of the three-dimensional current matrix;
[0033] Perform convolution of the first three-dimensional convolution kernel on the three-dimensional current matrix in the direction from the smallest layer to the largest layer of the three-dimensional current matrix with a step size of 1 to obtain a first variation feature.
[0034] Optionally, based on the first conduction feature and the second conduction feature, predicting the time point when a preset current value is reached to obtain a conduction time length, including:
[0035] Input the first conduction feature and the second conduction feature into a fully connected network to fuse the features to obtain a conduction feature;
[0036] Input the conduction feature and the preset current value into a prediction network to obtain a conduction time length; the conduction time length represents the time length that the contact of the switch can be used normally.
[0037] Optionally, based on the three-dimensional current matrix, obtaining the first straight line, the second straight line, the third straight line, and the fourth straight line, including:
[0038] Take the corner positions of the three-dimensional current matrix as the first corner position, the second corner position, the third corner position, the fourth corner position, the fifth corner position, the sixth corner position, the seventh corner position, and the eighth corner position;
[0039] The straight line passing through the first corner position, the three-dimensional center point and the eighth corner position is taken as the first straight line; the straight line passing through the second corner position, the three-dimensional center point and the seventh corner position is taken as the second straight line; the straight line passing through the third corner position, the three-dimensional center point and the sixth corner position is taken as the third straight line; the straight line passing through the fourth corner position, the three-dimensional center point and the fifth corner position is taken as the fourth straight line.
[0040] Optionally, the obtaining of the current matrices at multiple time points includes:
[0041] The rows of the current matrix represent the contact positions of the switch contacts to be detected;
[0042] The columns of the current matrix represent the values of the energizing current;
[0043] The values in the current matrix represent current differences; the current differences represent the differences between the current changed due to the change in resistance at the contact position and the energizing current.
[0044] In a second aspect, an embodiment of the present invention provides a conductive performance detection system for switch contacts, including:
[0045] An acquisition module, configured to acquire contact voltage drops and contact temperatures at multiple time points; the contact voltage drop represents the voltage loss when current passes through the switch contacts; the contact temperature represents the temperature rise when the switch contacts are in contact; acquire current matrices at multiple time points; the current matrices include multiple current differences; the current differences represent the differences between the currents changed due to the magnetic field under different energizing currents of the connected switch contacts;
[0046] A first conductive detection module, configured to construct a three-dimensional structure based on multiple current matrices, detect conductivity, and obtain a first conductive feature;
[0047] A second conductive detection module, configured to detect conductivity based on multiple current matrices, contact voltage drops and contact temperatures, and obtain a second conductive feature; the second conductive feature represents the characteristics of the conductivity change corresponding to current, voltage and temperature at multiple time points;
[0048] A prediction module, configured to predict the time point when a preset current value is reached based on the first conductive feature and the second conductive feature, and obtain a conductive time length.
[0049] Compared with the prior art, the embodiments of the present invention achieve the following beneficial effects:
[0050] The embodiments of the present invention also provide a method and a system for detecting the conductive performance of switch contacts.
[0051] In the present invention, by applying a certain energizing current across the contact tips and simultaneously collecting voltage, current, and temperature data of the contact positions of the contact tips for the switch in real time. An electric current matrix is constructed based on multiple contact positions and different energizing currents. A three-dimensional matrix is constructed based on the electric current matrices at multiple time points. It is possible to more accurately detect the changes in all current differences at different times, and the changes in current differences under different energizing currents and at different times. And the change in the resistance of the contact position of the contact tips for the switch is detected based on the current, voltage drop, and temperature. By combining the above two methods, it is possible to achieve the technical effect of more accurately evaluating the electrical conductivity of the contact tips through the current, voltage drop, and temperature at multiple time points. BRIEF DESCRIPTION OF THE DRAWINGS
[0052] Figure 1 FIG. is a flowchart of a method for detecting the electrical conductivity of a contact tip for a switch provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS
[0053] The present invention will be described in detail below with reference to the accompanying drawings.
[0054] Embodiment 1
[0055] As Figure 1 shown, an embodiment of the present invention provides a method for detecting the electrical conductivity of a contact tip for a switch, and the method includes:
[0056] S101: Obtain contact voltage drops and contact temperatures at multiple time points; the contact voltage drop represents the voltage loss when the energizing current passes through the contact tip for the switch; the contact temperature represents the temperature rise when the contact tip for the switch makes contact.
[0057] Wherein, the energizing current is the current value supplied by a direct current source across the contact tips for the switch.
[0058] Wherein, in this embodiment, the contact tip for the switch is a copper-chromium alloy self-supporting contact tip.
[0059] Wherein, the contact temperature is obtained because when the contact tip for the switch is connected, the temperature at the contact position rises, and the rising temperature affects the resistance.
[0060] Wherein, the contact position represents the position where the contact tip for the switch makes contact.
[0061] S102: Obtain current matrices at multiple time points; the current matrix represents the current changed by the magnetic field when different currents are input to the connected contact tips for the switch.
[0062] Wherein, the current matrix is a dot plot.
[0063] S103: Based on multiple current matrices, construct a three-dimensional structure, detect the electrical conductivity, and obtain a first electrical conductivity characteristic;
[0064] S104: Detect the conductivity based on multiple current matrices, contact voltage drops, and contact temperatures to obtain a second conductivity feature; the second conductivity feature represents the characteristics of conductivity changes corresponding to current, voltage, and temperature at multiple time points.
[0065] S105: Predict the time point when a preset current value is reached based on the first conductivity feature and the second conductivity feature to obtain the conduction time length.
[0066] Among them, the preset current value represents the current value at which the contact position of the switch contact is affected. In this embodiment, the current value that can be used normally is 0.23 amperes (A), and the preset current value is 0.22 amperes (A).
[0067] Optionally, the constructing a three-dimensional structure based on multiple current matrices, detecting the conductivity, and obtaining the first conductivity feature includes:
[0068] Overlay multiple current matrices to obtain a three-dimensional current matrix.
[0069] Among them, in this embodiment, the current matrix is q*p, where q represents the number of contact positions of the switch contact being detected. P represents different energizing currents. The number of current matrices obtained at multiple time points is k, then the number of rows of the three-dimensional current matrix is q, the number of columns is p, and the number of layers is k. q, p, and k are positive integers.
[0070] Based on the three-dimensional current matrix, detect the changes in all current differences at different times through a first convolutional network to obtain a first change feature.
[0071] Among them, the first convolutional network is a three-dimensional convolutional neural network (3D Convolutional Neural Networks, CNN).
[0072] The first convolutional network includes multiple first three-dimensional convolutional kernels;
[0073] Based on the three-dimensional current matrix, detect the changes in the current differences between different energizing currents and at different times through a second convolutional network to obtain a second change feature.
[0074] Among them, the second convolutional network is a three-dimensional convolutional neural network (3D Convolutional Neural Networks, CNN).
[0075] The second convolutional network includes multiple second three-dimensional convolutional kernels; the second convolutional network performs convolution along the diagonal;
[0076] Overlay the first change feature and the second change feature to obtain the first conductivity feature.
[0077] Among them, the first variation feature, the second variation feature, and the first conductivity feature are one-dimensional vectors.
[0078] Among them, the number of elements of the first conductivity feature is equal to the sum of the number of elements of the first variation feature and the number of elements of the second variation feature.
[0079] Optionally, detecting the conductivity based on multiple current matrices, contact voltage drops, and contact temperatures to obtain a second conductivity feature includes:
[0080] Summing multiple values of the current matrix to obtain a total current difference;
[0081] Constructing a conductivity relationship vector from the total current difference, contact temperature, and contact voltage drop; obtaining multiple conductivity relationship vectors corresponding to multiple time points.
[0082] Among them, the conductivity relationship vector contains 3 elements.
[0083] Inputting multiple conductivity relationship vectors into a temporal convolutional network in chronological order from early to late to obtain a second conductivity feature.
[0084] Among them, the temporal convolutional network (TCN).
[0085] Optionally, detecting the variation of the current difference between different energizing currents and at different times based on the three-dimensional current matrix through a second convolutional network to obtain a second variation feature includes:
[0086] Taking the center point of the three-dimensional current matrix as the three-dimensional center point.
[0087] Among them, the three-dimensional center point is at the position where the row is q / 2, the column is p / 2, and the layer is k / 2.
[0088] Based on the three-dimensional current matrix, obtaining the first straight line, the second straight line, the third straight line, and the fourth straight line;
[0089] The second three-dimensional convolution kernel is a 3*3*3 three-dimensional convolution kernel;
[0090] Based on the first straight line, the second straight line, the third straight line, and the fourth straight line, convolving the second three-dimensional convolution kernel with the three-dimensional current matrix to obtain a second variation feature.
[0091] Optionally, the convolving the second three-dimensional convolution kernel with the three-dimensional current matrix based on the first straight line, the second straight line, the third straight line, and the fourth straight line to obtain a second variation feature includes:
[0092] Obtaining the center point of the second three-dimensional convolution kernel as the center point of the second three-dimensional convolution kernel.
[0093] Among them, in this embodiment, the center point of the center point of the second 3D convolution kernel is the position where the row is 2, the column is 2, and the layer is 2 in the second 3D convolution kernel.
[0094] According to the order of the first straight line, the positions where the 3D current matrix coincides with the first straight line are sequentially marked to obtain a set of straight line positions.
[0095] Among them, the order of the first straight line is as follows: If a position where the 3D current matrix coincides with the first straight line is used as the first position, the position where the 3D current matrix coincides with the first straight line and is adjacent to the first position is used as the second position, and the position where the 3D current matrix coincides with the first straight line and is adjacent to the second position is used as the third position. The first position, the second position, and the third position are sequentially placed in the set of straight line positions.
[0096] Match the center point of the second 3D convolution kernel with the values in the set of straight line positions, and perform convolution on the 3D current matrix with a step size of 1 to obtain a second change feature; the second change feature is used to detect the feature of the common change of time and current.
[0097] Optionally, based on the 3D current matrix, by a first convolution network, detecting the change of all current differences at different times to obtain a first change feature, including:
[0098] The first 3D convolution kernel is a 3D convolution kernel of n*m*2; n represents the number of rows of the first 3D convolution kernel; m represents the number of columns of the first 3D convolution kernel; 2 corresponds to the number of layers of the 3D current matrix.
[0099] According to the direction of the layers of the 3D current matrix from small to large, with a step size of 1, perform convolution of the first 3D convolution kernel on the 3D current matrix to obtain a first change feature.
[0100] Optionally, based on the first conduction feature and the second conduction feature, predicting the time point to reach a certain current to obtain a conduction time length, including:
[0101] Input the first conduction feature and the second conduction feature into a fully connected network to fuse the features to obtain a conduction feature.
[0102] Among them, in this embodiment, it is a fully connected neural network (Full Connect Neural Network, FCN).
[0103] Input the conduction feature and a preset current value into a prediction network to obtain a conduction time length; the conduction time length represents the time length that the switch contact can be used normally.
[0104] Among them, the prediction network is a recurrent neural network (Recurrent Neural Network, RNN).
[0105] The prediction network is trained by using the current matrix, contact voltage drop and contact temperature at multiple historical time points and the marked conduction time length.
[0106] Among them, other labeled data such as labeled conductivity can also be used to train the prediction network to obtain a cluster of networks that can predict conductivity.
[0107] Optionally, the obtaining of eight first straight line position sets based on the three-dimensional current matrix includes:
[0108] Using the corner positions of the three-dimensional current matrix as the first corner position, the second corner position, the third corner position, the fourth corner position, the fifth corner position, the sixth corner position, the seventh corner position, and the eighth corner position;
[0109] A straight line passing through the first corner position, the three-dimensional center point, and the eighth corner position is defined as a first straight line; a straight line passing through the second corner position, the three-dimensional center point, and the seventh corner position is defined as a second straight line; a straight line passing through the third corner position, the three-dimensional center point, and the sixth corner position is defined as a third straight line; and a straight line passing through the fourth corner position, the three-dimensional center point, and the fifth corner position is defined as a fourth straight line.
[0110] Optionally, obtaining the current matrix at multiple time points includes:
[0111] The rows of the current matrix represent the detected contact positions of the switching contacts.
[0112] The columns of the current matrix represent the values of the energized current.
[0113] The values in the current matrix represent current differences; the current differences represent the difference between the current resulting from the resistance change at the contact location and the current being carried.
[0114] Among them, changes in the magnetic field at the contact position caused by temperature and current will cause changes in the resistance of the contact position.
[0115] Example 2
[0116] Based on the above-mentioned method for detecting the conductivity of a switch contact, an embodiment of the present invention further provides a system for detecting the conductivity of a switch contact, the system comprising:
[0117] An acquisition module for acquiring the contact voltage drop and contact temperature at multiple time points; the contact voltage drop represents the voltage loss that occurs when current passes through the switch contact; the contact temperature represents the temperature rise when the switch contacts are in contact; acquiring a current matrix at multiple time points; the current matrix includes a plurality of current differences; the current difference represents the difference in current changed by the magnetic field in the connected switch contacts under different energizing currents;
[0118] A first conductive detection module for constructing a three-dimensional structure based on a plurality of current matrices, detecting conductivity, and obtaining a first conductive feature;
[0119] A second conductive detection module for detecting conductivity based on a plurality of current matrices, contact voltage drop, and contact temperature, and obtaining a second conductive feature; the second conductive feature represents the characteristics of the change in conductivity corresponding to current, voltage, and temperature at multiple time points;
[0120] A prediction module for predicting the time point when a preset current value is reached based on the first conductive feature and the second conductive feature, and obtaining the conductive time length.
[0121] In the specification provided herein, a large number of specific details are set forth. It will be understood, however, that embodiments of the invention may be practiced without these specific details. In some instances, well-known methods, structures and techniques have not been shown in detail in order not to obscure an understanding of this description.
[0122] Each component embodiment of the present invention may be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. Those skilled in the art should understand that a microprocessor or a digital signal processor (DSP) may be used in practice to implement some or all of the functions of some or all of the components in the device according to the embodiments of the present invention. The present invention may also be implemented as a device or device program for executing part or all of the methods described herein (e.g., a computer program and a computer program product). Such a program for implementing the present invention may be stored on a computer-readable medium, or may be in the form of one or more signals. Such signals may be downloaded from an Internet website, or provided on a carrier signal, or in any other form.
Claims
1. A method for detecting the electrical conductivity of a contact for a switch, characterized in that, Including: Obtaining contact voltage drops and contact temperatures at multiple time points; the contact voltage drop represents the voltage loss when the energized current passes through the switch contact; The contact temperature represents the temperature rise when the switch contacts; Obtaining current matrices at multiple time points; the current matrix includes multiple current differences; Based on multiple current matrices, constructing a three-dimensional structure, detecting conductivity, and obtaining a first conductivity feature; Based on multiple current matrices, contact voltage drops, and contact temperatures, detecting conductivity, and obtaining a second conductivity feature; the second conductivity feature represents the characteristics of the conductivity change corresponding to current, voltage, and temperature at multiple time points; Based on the first conductivity feature and the second conductivity feature, predicting the time point when the preset current value is reached, and obtaining the conductive time length.
2. The method for detecting the electrical conductivity of a contact for a switch according to claim 1, characterized in that The step of based on multiple current matrices, constructing a three-dimensional structure, detecting conductivity, and obtaining a first conductivity feature includes: Stacking multiple current matrices to obtain a three-dimensional current matrix; Based on the three-dimensional current matrix, through a first convolutional network, detecting the changes of all current differences at different times, and obtaining a first change feature; The first convolutional network includes multiple first three-dimensional convolutional kernels; Based on the three-dimensional current matrix, through a second convolutional network, detecting the changes of current differences under different energized currents and different times, and obtaining a second change feature; The second convolutional network includes multiple second three-dimensional convolutional kernels; the second convolutional network performs convolution along the diagonal; Superimposing the first change feature and the second change feature to obtain a first conductivity feature; Wherein, the first change feature, the second change feature, and the first conductivity feature are one-dimensional vectors.
3. The method for detecting the electrical conductivity of a contact for a switch according to claim 1, characterized in that The step of based on multiple current matrices, contact voltage drops, and contact temperatures, detecting conductivity, and obtaining a second conductivity feature includes: Summing multiple values of the current matrix to obtain a total current difference; Constructing a conductivity relationship vector from the total current difference, contact temperature, and contact voltage drop; multiple conductivity relationship vectors are obtained corresponding to multiple time points; Inputting multiple conductivity relationship vectors into a temporal convolutional network in chronological order from early to late to obtain a second conductivity feature.
4. A method for detecting the electrical conductivity of a contact for a switch according to claim 2, characterized in that, The step of based on the three-dimensional current matrix, through a second convolutional network, detecting the changes of current differences under different energized currents and different times, and obtaining a second change feature includes: Taking the center point of the three-dimensional current matrix as the three-dimensional center point; Based on the three-dimensional current matrix, obtaining a first straight line, a second straight line, a third straight line, and a fourth straight line; The second three-dimensional convolutional kernel is a 3*3*3 three-dimensional convolutional kernel; Based on the first straight line, the second straight line, the third straight line, and the fourth straight line, convolving the second three-dimensional convolutional kernel with the three-dimensional current matrix to obtain a second change feature.
5. The method for detecting the electrical conductivity of a contact for a switch according to claim 4, characterized in that, The step of based on the first straight line, the second straight line, the third straight line, and the fourth straight line, convolving the second three-dimensional convolutional kernel with the three-dimensional current matrix to obtain a second change feature includes: Obtaining the center point of the second three-dimensional convolutional kernel as the second three-dimensional convolutional kernel center point; Marking the positions where the three-dimensional current matrix coincides with the first straight line in sequence according to the order of the first straight line to obtain a straight line position set; Match the center point of the second three-dimensional convolution kernel with the values in the set of straight line positions, and perform convolution on the three-dimensional current matrix with a step size of 1 to obtain a second variation feature; the second variation feature is used to detect the features that change jointly with time and current.
6. A method for detecting the electrical conductivity of a contact for a switch according to claim 2, characterized in that, Based on the three-dimensional current matrix, through a first convolution network, detect the changes in all current differences at different times to obtain a first variation feature, including: The first three-dimensional convolution kernel is a three-dimensional convolution kernel of n*m*2; n represents the number of rows of the first three-dimensional convolution kernel; m represents the number of columns of the first three-dimensional convolution kernel; 2 corresponds to the number of layers of the three-dimensional current matrix; In the direction from the smallest layer to the largest layer of the three-dimensional current matrix, perform convolution of the first three-dimensional convolution kernel on the three-dimensional current matrix with a step size of 1 to obtain a first variation feature.
7. A method for detecting the electrical conductivity of a contact for a switch according to claim 1, characterized in that Based on the first conductivity feature and the second conductivity feature, predict the time point when the preset current value is reached to obtain the conduction time length, including: Input the first conductivity feature and the second conductivity feature into a fully connected network to fuse the features to obtain a conductivity feature; Input the conductivity feature and the preset current value into a prediction network to obtain the conduction time length; the conduction time length represents the time length during which the contact of the switch can be used normally.
8. A method for detecting the electrical conductivity of a contact for a switch according to claim 4, characterized in that, Based on the three-dimensional current matrix, obtain a first straight line, a second straight line, a third straight line, and a fourth straight line, including: Regard the corner positions of the three-dimensional current matrix as the first corner position, the second corner position, the third corner position, the fourth corner position, the fifth corner position, the sixth corner position, the seventh corner position, and the eighth corner position; Regard the straight line passing through the first corner position, the three-dimensional center point, and the eighth corner position as the first straight line; regard the straight line passing through the second corner position, the three-dimensional center point, and the seventh corner position as the second straight line; regard the straight line passing through the third corner position, the three-dimensional center point, and the sixth corner position as the third straight line; regard the straight line passing through the fourth corner position, the three-dimensional center point, and the fifth corner position as the fourth straight line.
9. A method for detecting the electrical conductivity of a contact for a switch according to claim 1, characterized in that, Obtain the current matrices at multiple time points, including: The rows of the current matrix represent the contact positions of the contacts of the switch being detected; The columns of the current matrix represent the values of the energized current; The values in the current matrix represent the current difference; the current difference represents the difference between the current changed due to the resistance at the contact position and the energized current.
10. A conductive performance detection system for a contact of a switch, characterized in that, Include: An acquisition module for acquiring the contact voltage drop and the contact temperature at multiple time points; the contact voltage drop represents the voltage loss when the current passes through the contact of the switch; The contact temperature represents the temperature rise when the contacts of the switch are in contact; acquire the current matrices at multiple time points; the current matrix includes multiple current differences; the current difference represents the difference between the currents changed due to the magnetic field under different energized currents of the connected contacts of the switch; A first conductivity detection module for constructing a three-dimensional structure based on multiple current matrices, detecting conductivity, and obtaining a first conductivity feature; A second conductivity detection module for detecting conductivity based on multiple current matrices, the contact voltage drop, and the contact temperature to obtain a second conductivity feature; the second conductivity feature represents the feature of the conductivity change corresponding to the current, voltage, and temperature at multiple time points; A prediction module, configured to predict a time point when a preset current value is reached based on a first conductive feature and a second conductive feature, so as to obtain a conductive time length.