Method and device for determining electrodynamic force, electronic device, and storage medium

Abstract

The present invention discloses a method and device for determining electrodynamic force, an electronic device, and a storage medium. The method includes: obtaining first parameter information of a first switch in a first state; inputting the current density of the first switch, the magnetic induction intensity of the first switch, the conductivity of the moving blade of the first switch, the electric field intensity of the first switch, and the volume of the first switch into a first target model to obtain a first Ampere force of the first switch; inputting the magnetic permeability of the moving blade of the first switch, the magnetic induction intensity of the first switch, the contact area between the moving blade and the static contact of the first switch, the density of the moving blade of the first switch, the current density of the first switch, and the electric field intensity of the first switch into a second target model to obtain a first Holm force of the first switch; and determining a first electrodynamic force of the first switch according to the first Ampere force and the first Holm force. Through the first target model and the second target model, the magnitude of the first electrodynamic force can be obtained in real time, quickly, and accurately finally.

Description

Technical Field

[0001] The present invention relates to the technical fields of power electronics and electromagnetic field modeling, and particularly to a method and device for determining electrodynamic force, an electronic device, and a storage medium. Background Art

[0002] In the fields of power electronics and electromagnetic field modeling, it is of great significance to accurately obtain relevant force parameters such as electrodynamic force that characterize the health status of switches. Traditional methods for obtaining the electrodynamic force of switches usually rely on empirical formulas and manual calculations. However, these methods have many limitations and defects. For example, low calculation accuracy, high complexity, inability to obtain dynamically changing data in real time, etc. Summary of the Invention

[0003] The present invention provides a method and device for determining the electrodynamic force of a switch, an electronic device, and a storage medium to solve the problems of low accuracy and high complexity in calculating the electrodynamic force of the switch.

[0004] According to a first aspect of the present invention, there is provided a method for determining the electrodynamic force of a switch, where the switch includes a moving blade and at least one stationary contact, and the method for determining the electrodynamic force of the switch includes:

[0005] Obtain first parameter information of a first switch in a first state, where the first parameter information of the first switch in the first state includes: the magnetic induction intensity of the first switch, the electric field intensity of the first switch, the volume of the first switch, the current density of the first switch, the conductivity of the moving blade of the first switch, the magnetic permeability of the moving blade of the first switch, the density of the moving blade of the first switch, and the contact area between the moving blade and the stationary contact of the first switch;

[0006] Input the current density of the first switch, the magnetic induction intensity of the first switch, the conductivity of the moving blade of the first switch, the electric field intensity of the first switch, and the volume of the first switch into a first target model to obtain a first Ampere force of the first switch;

[0007] Input the magnetic permeability of the moving blade of the first switch, the magnetic induction intensity of the first switch, the contact area between the moving blade and the stationary contact of the first switch, the density of the moving blade of the first switch, the current density of the first switch, and the electric field intensity of the first switch into a second target model to obtain a first Holm force of the first switch;

[0008] Determine a first electrodynamic force of the first switch according to the first Ampere force and the first Holm force.

[0009] Optionally, the determining the first electrodynamic force of the first switch according to the first Ampere force and the first Holm force includes:

[0010] The sum of the product of the first Ampere force multiplied by the first weight and the product of the first Holm force multiplied by the second weight is determined as the first electrodynamic force of the first switch.

[0011] Optionally, before obtaining the first parameter information of the first switch in the first state, it further includes: obtaining a set of second parameter information for each of the c second switches in each of the m second states, as well as the actual Ampere force and the actual Holm force corresponding to each set of second parameter information; the second parameter information of the second switch in the second state includes: the magnetic induction intensity of the second switch, the electric field intensity of the second switch, the volume of the second switch, the current density of the second switch, the conductivity of the moving blade of the second switch, the magnetic permeability of the moving blade of the second switch, the density of the moving blade of the second switch, and the contact area between the moving blade and the static contact of the second switch; m, a, and c are all integers greater than or equal to 1.

[0012] The current density of the second switch, the magnetic induction intensity of the second switch, the conductivity of the moving blade of the second switch, the electric field intensity of the second switch, the volume of the second switch, and the corresponding actual Ampere force are used as the first training data set.

[0013] The magnetic permeability of the moving blade of the second switch, the magnetic induction intensity of the second switch, the contact area between the moving blade and the static contact of the second switch, the density of the moving blade of the second switch, the current density of the second switch, the electric field intensity of the second switch, and the actual Holm force are used as the second training data set.

[0014] Input the c·m·a first training data sets into the first initial model to sequentially obtain c·m·a first initial Ampere forces; input the c·m·a second training data sets into the second initial model to sequentially obtain c·m·a first initial Holm forces.

[0015] Adjust the first weight according to the first initial Ampere force and the actual Ampere force; adjust the second weight according to the first initial Holm force and the actual Holm force.

[0016] Optionally, the adjusting the first weight according to the first initial Ampere force and the actual Ampere force includes:

[0017] Taking the ratio of the p-th actual Ampere force to the first initial Ampere force obtained by inputting the p-th first training data set into the first initial model as the first weight corresponding to the p-th training of the first initial model, and determining the first loss function value corresponding to the first weight at the p-th training; p is an integer greater than or equal to 1 and less than or equal to c·m·a.

[0018] Input the second parameter information into the first policy network to obtain a first probability distribution, where the first probability distribution is used to represent the adjustment direction of the model parameters of the first initial model;

[0019] Input the second parameter information into the first value network to obtain a first value, where the first value is used to identify the adjustment value of the first policy network;

[0020] The adjustment of the second weight according to the first initial Holm force and the actual Holm force includes:

[0021] Use the ratio of the p-th actual Holm force to the first initial Holm force obtained by inputting the p-th second training data set into the second initial model as the second weight corresponding to the second initial model during the p-th training, and determine the second loss function value corresponding to the second weight during the p-th training;

[0022] Input the second parameter information into the second policy network to obtain a second probability distribution, where the second probability distribution is used to represent the adjustment direction of the model parameters of the second initial model;

[0023] Input the second parameter information into the second value network to obtain a second value, where the second value is used to identify the adjustment value of the second policy network.

[0024] Optionally, the first parameter information of the first switch in the first state includes the electric field strength of the first switch;

[0025] The obtaining of the first parameter information of the first switch in the first state includes:

[0026] Obtain the electric field distribution image of the first switch in the first state;

[0027] Determine the electric field strength at the center position of the moving blade of the first switch according to the electric field distribution image, and use the electric field strength at the center position of the moving blade of the first switch as the electric field strength of the first switch.

[0028] Optionally, the first parameter information of the first switch in the first state includes the current density of the first switch;

[0029] The obtaining of the first parameter information of the first switch in the first state includes:

[0030] Obtain the current values of the first switch at different times respectively;

[0031] Generate an initial current waveform function according to the current values of the first switch at different times;

[0032] Generate a predicted current waveform function based on the ensemble learning algorithm and the initial current waveform function;

[0033] Determine the maximum value of the predicted current according to the predicted current waveform function, and determine the current density of the first switch according to the maximum value of the predicted current.

[0034] Optionally, the first parameter information of the first switch in the first state includes the density of the moving blade of the first switch, the conductivity of the moving blade of the first switch, and the magnetic permeability of the moving blade of the first switch;

[0035] The obtaining of the first parameter information of the first switch in the first state includes:

[0036] Based on the material of the moving blade of the first switch and a preset corresponding relationship, determine the density of the moving blade of the first switch, the conductivity of the moving blade of the first switch, and the magnetic permeability of the moving blade of the first switch; wherein, the preset corresponding relationship is the corresponding relationship between the material and the density, conductivity, and magnetic permeability.

[0037] According to a second aspect of the present invention, there is provided a device for determining the electrodynamic force of a switch, the switch including a moving blade and at least one stationary contact, and the device for determining the electrodynamic force of the switch includes:

[0038] A parameter information acquisition module, configured to acquire first parameter information of a first switch in a first state, where the first parameter information of the first switch in the first state includes: the magnetic induction intensity of the first switch, the electric field intensity of the first switch, the volume of the first switch, the current density of the first switch, the conductivity of the moving blade of the first switch, the magnetic permeability of the moving blade of the first switch, the density of the moving blade of the first switch, and the contact area between the moving blade and the stationary contact of the first switch. A first Ampere force generation module, configured to input the current density of the first switch, the magnetic induction intensity of the first switch, the conductivity of the moving blade of the first switch, the electric field intensity of the first switch, and the volume of the first switch into a first target model to obtain the first Ampere force of the first switch;

[0039] A first Holm force generation module, configured to input the magnetic permeability of the moving blade of the first switch, the magnetic induction intensity of the first switch, the contact area between the moving blade and the stationary contact of the first switch, the density of the moving blade of the first switch, the current density of the first switch, and the electric field intensity of the first switch into a second target model to obtain the first Holm force of the first switch;

[0040] An electrodynamic force generation module, configured to determine the first electrodynamic force of the first switch according to the first Ampere force and the first Holm force.

[0041] According to a third aspect of the present invention, there is provided an electronic device, the electronic device including:

[0042] at least one processor; and

[0043] a memory communicatively connected to the at least one processor; wherein

[0044] the memory stores a computer program executable by the at least one processor, and when the computer program is executed by the at least one processor, the at least one processor is enabled to execute the method for determining the electrodynamic force of the switch according to any one of the first aspect.

[0045] According to a fourth aspect of the present invention, there is provided a computer-readable storage medium storing computer instructions for causing a processor to implement the method for determining the electrodynamic force of the switch according to the first aspect when executed.

[0046] In the technical solution of the embodiment of the present invention, after obtaining the first parameter information of the first switch in the first state, the part of the first parameter information related to the Ampere force in the first parameter information is input into the trained first target model to obtain the first Ampere force, and the part of the first parameter information related to the Holm force in the first parameter information is input into the trained second target model to obtain the first Holm force, and then the first electrodynamic force is obtained according to the first Ampere force and the first Holm force. Since the first target model and the second target model have been trained in advance, the output values of the models are the first Ampere force and the first Holm force, and the first Ampere force and the first Holm force of the first switch can be accurately obtained, and then the first electrodynamic force can be obtained. Through the first target model and the second target model, the magnitude of the first electrodynamic force of the first switch can be obtained in real time, quickly and accurately, solving the problems of low calculation accuracy and high complexity of the traditional method.

[0047] It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of the present invention, nor is it used to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention, and those of ordinary skill in the art can obtain other drawings without creative efforts based on these drawings.

[0049] Figure 1 is a flowchart of a method for determining the electrodynamic force of a switch provided by an embodiment of the present invention;

[0050] Figure 2Flow chart of another method for determining the electrodynamic force of a switch provided by an embodiment of the present invention;

[0051] Figure 3 Comparison diagram of an actual current waveform and a predicted current waveform provided by an embodiment of the present invention;

[0052] Figure 4 Flow chart of another method for determining the electrodynamic force of a switch provided by an embodiment of the present invention;

[0053] Figure 5 Network architecture diagram of a reinforcement learning provided by an embodiment of the present invention;

[0054] Figure 6 Structural schematic diagram of a device for determining an electrodynamic force provided by an embodiment of the present invention;

[0055] Figure 7 Structural schematic diagram of an electronic device for implementing the method for determining an electrodynamic force according to an embodiment of the present invention. Detailed implementation manners

[0056] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0057] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above accompanying drawings are used to distinguish similar objects, and do not necessarily need to describe a specific order or sequence. It should be understood that such used data may be interchanged under appropriate circumstances so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0058] Figure 1 Flow chart of a method for determining the electrodynamic force of a switch provided by an embodiment of the present invention. This embodiment is applicable to the situation of determining the health status of a switch according to the magnitude of the electrodynamic force of the switch. The switch includes a moving blade and at least one static contact. The switch can be a three-position switch and a load switch, such as Figure 1 shown, the method includes:

[0059] S110: Obtain the first parameter information of the first switch in the first state.

[0060] The first state of the first switch can be understood as the state of the first switch at the current moment. The first state of the first switch includes the state where the moving blade of the first switch and all the static contacts of the first switch are not in contact, or the state where the moving blade of the first switch and any one of the static contacts of the first switch are in contact.

[0061] Exemplarily, the switch includes a moving blade and two static contacts, and the two static contacts are the upper static contact and the earthing knife respectively. The first switch is any switch that needs to determine the health status according to the electrodynamic force, that is, any switch to be detected for the electrodynamic force. The first state of the first switch includes two closed states and one open state. In the open state, the moving blade of the first switch and any one of the static contacts of the first switch are not in contact. The two closed states include the first closed state and the second closed state. In the first closed state, the moving blade of the first switch is in contact with the upper static contact of the first switch. In the second closed state, the moving blade of the first switch is in contact with the earthing knife of the first switch. The current flowing through the first switch is different in different closed states. The specific numerical values of at least some of the first parameter information corresponding to the first switch are different in different first states.

[0062] The first parameter information of the first switch in the first state includes: the magnetic induction intensity of the first switch, the electric field intensity of the first switch, the volume of the first switch, the current density of the first switch, the conductivity of the moving blade of the first switch, the magnetic permeability of the moving blade of the first switch, the density of the moving blade of the first switch, and the contact area between the moving blade and the static contact of the first switch.

[0063] S120: Input the current density of the first switch, the magnetic induction intensity of the first switch, the conductivity of the moving blade of the first switch, the electric field intensity of the first switch, and the volume of the first switch into the first target model to obtain the first Ampere force of the first switch.

[0064] The first Ampere force is the force exerted on the energized first switch in the magnetic field. Among them, when there is current flowing through the first switch, the current will generate a magnetic field, making the first switch in the magnetic field.

[0065] The current density of the first switch, the magnetic induction intensity of the first switch, the conductivity of the moving blade of the first switch, the electric field strength of the first switch, and the volume of the first switch are all related to the magnitude of the Ampere force of the first switch. The first target model is a pre-trained model. Input the current density of the first switch, the magnetic induction intensity of the first switch, the conductivity of the moving blade of the first switch, the electric field strength of the first switch, and the volume of the first switch into the first target model, and the output value of the first target model is the first Ampere force of the first switch.

[0066] S130: Input the magnetic permeability of the moving blade of the first switch, the magnetic induction intensity of the first switch, the contact area between the moving blade and the static contact of the first switch, the density of the moving blade of the first switch, the current density of the first switch, and the electric field strength of the first switch into the second target model to obtain the first Holm force of the first switch.

[0067] Since the contact surface between the moving blade and the static contact is uneven, when current flows through the uneven contact surface, the current contracts near the contact surface, generating a contracting electro-dynamic repulsive force between the moving blade and the static contact, that is, the Holm force. Therefore, when current flows through the first switch, the first Holm force of the first switch in the first state will be generated.

[0068] The magnetic permeability of the moving blade of the first switch, the magnetic induction intensity of the first switch, the contact area between the moving blade and the static contact of the first switch, the density of the moving blade of the first switch, the current density of the first switch, and the electric field strength of the first switch are all related to the magnitude of the Holm force of the first switch. The second target model is a pre-trained model. Input the magnetic permeability of the moving blade of the second switch, the magnetic induction intensity of the first switch, the contact area between the moving blade and the static contact of the first switch, the density of the moving blade of the first switch, the current density of the first switch, and the electric field strength of the first switch into the second target model, and the output value of the second target model is the first Holm force of the first switch.

[0069] S140: Determine the first electromotive force of the first switch according to the first Ampere force and the first Holm force.

[0070] The first electromotive force is related to both the magnitude of the first Ampere force and the magnitude of the first Holm force, and is a comprehensive representation of the first Ampere force and the first Holm force. Determine the health status of the first switch in the first state according to the magnitude of the first electromotive force.

[0071] In the technical solution of the embodiment of the present invention, after obtaining the first parameter information of the first switch in the first state, the part of the first parameter information related to the Ampere force in the first parameter information is input into the trained first target model to obtain the first Ampere force, and the part of the first parameter information related to the Holm force in the first parameter information is input into the trained second target model to obtain the first Holm force. Then, the first electromotive force is obtained according to the first Ampere force and the first Holm force. Since the first target model and the second target model have been trained in advance, the output values of the models are the first Ampere force and the first Holm force, so the first Ampere force and the first Holm force of the first switch can be accurately obtained, and then the first electromotive force can be obtained. Through the first target model and the second target model, the magnitude of the first electromotive force of the first switch can be obtained in real time, quickly and accurately, solving the problems of low calculation accuracy and high complexity of the traditional method.

[0072] Figure 2 It is a flowchart of another method for determining the electromotive force of a switch provided by an embodiment of the present invention. Figure 2 For Figure 1 the further refinement of S110 in Figure 2 , the method includes:

[0073] S111: Obtain the electric field distribution image of the first switch in the first state; wherein, the first parameter information of the first switch in the first state includes the electric field strength of the first switch.

[0074] Capture the electric field distribution image of the first switch in the first state through a spatial electromagnetic field visualizer. In the electric field distribution image, the electric field strengths at different positions of the first switch, such as different positions of the moving blade and different positions of the static contact, are different, so the colors at different positions of the first switch are different. In the electric field distribution image, the colors in the image correspond one-to-one with the magnitudes of the electric field strengths.

[0075] S121: Determine the electric field strength at the center position of the moving blade of the first switch according to the electric field distribution image, and use the electric field strength at the center position of the moving blade of the first switch as the electric field strength of the first switch.

[0076] Obtain the color at the center position of the moving blade from the electric field distribution image, and then determine the magnitude of the electric field strength at the center position of the moving blade of the first switch through the correspondence between the color and the electric field strength. The electric field strength of the first switch is characterized by the electric field strength at the center position of the moving blade of the first switch, and then the electric field strength of the first switch is obtained.

[0077] S131: Obtain the current values of the first switch at different times respectively.

[0078] When the current of the first switch meets the conditions of the rated peak withstand current and the rated short-circuit withstand current, the magnitude of the current in the first switch is obtained through a current transformer at different times, and then the current values in the first switch at different times are obtained.

[0079] S141: Generate an initial current waveform function according to the current values of the first switch at different times.

[0080] The current values of the first switch obtained at different times are fitted to obtain an initial current waveform function , specifically:

[0081] ; where represents the maximum value of the actual current, represents the angular frequency of the actual current, represents the initial phase angle of the actual current.

[0082] S151: Generate a predicted current waveform function based on the integrated learning algorithm and the initial current waveform function.

[0083] Use the integrated learning algorithm to integrate and analyze the initial current waveform function, and predict the change trend of the current:

[0084] ;

[0085] where represents the predicted function, represents the weight of the f-th weak learner in the integrated learning model, represents the prediction result of the f-th weak learner, and F represents the number of weak learners. By weighted averaging the prediction results of multiple weak learners, the overall prediction accuracy and stability can be improved.

[0086] Predicted current waveform function Satisfy:

[0087] ; where represents the maximum value of the predicted current, represents the angular frequency of the predicted current, represents the initial phase angle of the predicted current.

[0088] Figure 3 This is a comparison chart of the actual current waveform and the predicted current waveform provided by the embodiments of the present invention. Figure 3The abscissa in the figure is time, with the unit of seconds, and the ordinate is current, with the unit of amperes. It can be seen from the figure that the actual current waveform (blue curve) has certain fluctuations at different time points, reflecting the current changes under different load and short - circuit conditions. The predicted current waveform (orange curve), on the other hand, smooths out these fluctuations through the moving average method, better showing the trend of current change.

[0089] Specifically, the orange curve in the figure closely follows the overall trend of the blue curve, indicating that by integrating and analyzing sensor data through the ensemble learning algorithm, the trend of current change can be accurately predicted. This trend prediction helps to better understand the dynamic behavior of the current and provides strong support for the optimization of load switches and fault diagnosis. By comparing the actual waveform and the predicted waveform, the effectiveness of the ensemble learning algorithm in capturing the overall trend can be observed, and at the same time, the noise and short - term fluctuations in the waveform can be identified. Using the weighted average of the prediction results of multiple weak learners improves the overall prediction accuracy and stability, making up for the insensitivity of traditional methods to changes in complex electromagnetic environments.

[0090] S161: Determine the maximum value of the predicted current according to the predicted current waveform function, and determine the current density of the first switch according to the maximum value of the predicted current.

[0091] After obtaining the predicted current waveform function, the maximum value of the predicted current can be obtained. Take the ratio of the maximum value of the predicted current to the cross - sectional area of the first switch in the direction perpendicular to the actual current direction as the current density of the first switch.

[0092] S171: Based on the material of the moving blade of the first switch and the preset corresponding relationship, determine the density, conductivity, and permeability of the moving blade of the first switch; where the preset corresponding relationship is the corresponding relationship between the material and density, conductivity, and permeability.

[0093] Use the BERT model to process the literature and databases related to material properties, and automatically extract and summarize the conductivity, permeability, and density characteristics of the materials. Among them, the preset corresponding relationship can be a preset corresponding table. Table 1 shows the conductivity, permeability, and density data of several materials extracted from the literature and databases by the BERT model in this embodiment.

[0094] Table 1: Corresponding relationship table between different materials of the switch and conductivity, permeability, and density

[0095]

[0096] As can be seen from the above table, silver has the highest conductivity of 6.30 × 10^7 S / m and is suitable for high-performance conductive applications. Copper has a conductivity of 5.96 × 10^7 S / m and is widely used in wires and cables. Aluminum has a conductivity of 3.77 × 10^7 S / m and the lowest density of 2.70 g / cm³, making it suitable for lightweight applications. Although gold has a slightly lower conductivity of 4.10 × 10^7 S / m, it is often used in high-end electronic devices due to its corrosion resistance. Iron has a conductivity of 1.0 × 10^7 S / m, but its permeability is the highest at 6.3 × 10^-3 H / m, making it suitable for use as a magnetic material. The characteristics of each material provide a scientific basis for different applications and help in selecting the best material.

[0097] After determining the material of the moving blade of the first switch, the density, permeability, and conductivity of the corresponding moving blade are determined by looking up the table.

[0098] In an alternative method, when obtaining the contact area between the first switch and any static contact, a contact liquid is coated on the moving blade of the first switch. The first switch is controlled to close for a preset duration and then opened, and then the area of the remaining contact liquid on the moving blade of the first switch is measured as the contact area between the moving blade and the static contact of the first switch. Here, when the first switch is in a state where the moving blade is not in contact with all static contacts, the contact area between the moving blade and the static contact of the first switch is regarded as 0.

[0099] Optionally, the magnetic induction intensity B of the first switch is obtained through the formula:

[0100] ; where is the permeability of the moving blade of the first switch, I is the effective value of the current flowing through the first switch, and k is the length of the first switch in the direction perpendicular to the current direction.

[0101] Optionally, before inputting the first parameter information of the first switch in the first state into the first target model and the second target model, each piece of first parameter information can be standardized. Standardization refers to converting data with different dimensions to the same dimension so that they have the same measurement standard.

[0102] Optionally, the volume of the first switch is the volume of the overall structure composed of the moving blade and all static contacts of the first switch. The volume of the first switch is determined before the switch leaves the factory, and the volume of the first switch when it leaves the factory can be directly obtained later.

[0103] S181: Input the current density, magnetic induction intensity, conductivity of the moving blade, electric field intensity, and volume of the first switch into the first target model to obtain the first Ampere force of the first switch.

[0104] S191: Input the magnetic permeability of the moving blade of the first switch, the magnetic induction intensity of the first switch, the contact area between the moving blade and the static contact of the first switch, the density of the moving blade of the first switch, the current density of the first switch, and the electric field intensity of the first switch into the second target model to obtain the first Holm force of the first switch.

[0105] S201: Determine the first electrodynamic force of the first switch according to the first Ampere force and the first Holm force.

[0106] Figure 4 It is a flowchart of another method for determining the electrodynamic force of a switch provided by an embodiment of the present invention. Refer to Figure 4 and the method includes:

[0107] S112: Obtain a set of second parameter information for each of the c second switches in each of the m second states, as well as the actual Ampere force and the actual Holm force corresponding to each set of second parameter information, where m, a, and c are all integers greater than or equal to 1.

[0108] Among them, the second parameter information of the second switch in the second state includes: the magnetic induction intensity of the second switch, the electric field intensity of the second switch, the volume of the second switch, the current density of the second switch, the conductivity of the moving blade of the second switch, the magnetic permeability of the moving blade of the second switch, the density of the moving blade of the second switch, and the contact area between the moving blade and the static contact of the second switch.

[0109] According to the calculation formula of the Ampere force and the calculation formula of the Holm force, the parameter information related to the Ampere force and the parameter information related to the Holm force can be determined. Specifically:

[0110] ;

[0111] ;

[0112] Among them, represents the calculation model of the Ampere force, represents the calculation model of the Holm force, represents the current density of the switch, B represents the magnetic induction intensity of the switch, σ represents the conductivity of the moving blade, E represents the electric field intensity of the switch, represents the vacuum permittivity, V represents the volume of the switch, H represents the magnetic field intensity of the switch, represents the curl of the magnetic field, A represents the contact area between the moving blade and the static contact, ρ represents the density of the moving blade, , represents the magnetic permeability of the moving blade, where .

[0113] The second switch is any one of the switches having the same structure as the first switch, that is, the second switch is any one of the switches including a moving blade and a static contact. All c second switches can be switches of the same batch as the first switch, so that the performance of the second switch is similar to that of the first switch, thereby improving the accuracy of training the first initial model and the second initial model based on the second parameter information obtained from the second switch. Alternatively, the c second switches may include the first switch, and this embodiment does not make specific limitations thereon.

[0114] The second state includes the state in which the moving blade of the second switch does not contact all the static contacts of the second switch and the state in which the moving blade of the second switch contacts each static contact of the second switch. The second state of the second switch includes two closed states and one open state. In the open state, the moving blade of the second switch does not contact any static contact of the second switch. The two closed states include a third closed state and a fourth closed state. In the third closed state, the moving blade of the second switch contacts the upper static contact of the second switch. In the fourth closed state, the moving blade of the second switch contacts the earthing knife of the second switch. The current flowing through the second switch is different in different closed states. When the second switch is in different second states, the specific numerical values of at least some of the second parameter information corresponding thereto are different. Optionally, m is equal to 3, and a set of second parameter information in each of the 3 second states of each second switch is obtained. The 3 states are: the state in which the moving blade of the second switch does not contact any static contact of the second switch, the state in which the moving blade of the second switch contacts the upper static contact of the second switch, and the state in which the moving blade of the second switch contacts the earthing knife of the second switch. In each state of each second switch, the magnitude of the current input to the second switch is changed by changing the current source of the second switch, thereby obtaining a set of second parameter information.

[0115] In an alternative embodiment, the materials of the moving blades of each of the c second switches are different. Optionally, c = 5. Then, multiple sets of second parameter information of the five second switches in each second state are obtained. The materials of the moving blades of the five second switches are different. The materials of the moving blades of the five second switches can be copper, iron, aluminum, silver, and gold respectively. For each second state of each second switch, at least three sets of second parameter information are obtained respectively. Specifically, for the same second switch, the current source supplying current to the second switch is changed, and the second parameter information of the second switch under each current source is obtained. Among them, under the same second switch, the electric field intensity, magnetic induction intensity, and current density of the second switch corresponding to different current sources are different, while the contact area between the moving blade and the static contact of the second switch is the same, the density of the moving blade of the second switch is the same, the conductivity of the moving blade of the second switch is the same, and the magnetic permeability of the moving blade of the second switch is the same. When the materials of the moving blades of different second switches are different in the same second state, the density, conductivity, and magnetic permeability of the moving blades of the second switches are different, and the contact area between the moving blade and the static contact of the second switch is different.

[0116] Among them, the method for obtaining the magnetic induction intensity of the second switch is the same as that of the first switch, the method for obtaining the electric field intensity of the second switch is the same as that of the first switch described in S111 - S121, the method for obtaining the volume of the second switch is the same as that of the first switch, the method for obtaining the current density of the second switch is the same as that of the first switch described in S131 - 161, the conductivity, magnetic permeability, and density of the moving blade of the second switch are the same as those of the moving blade of the first switch described in S171, and the contact area between the moving blade and the static contact of the second switch is the same as that of the first switch described above. Details are not described herein again.

[0117] The actual Ampere force and the actual Holm force are real data directly obtained by experiments or measurements. Optionally, high-precision instruments such as stress sensors and electromagnetic field sensors are used to measure the Ampere force of the second switch as the actual Ampere force of the second switch and measure the Holm force of the second switch as the actual Holm force of the second switch.

[0118] In this embodiment, a sets of second parameter information are obtained for each state of each second switch in three second states, ensuring that the second parameter information of the second switch under different working conditions is covered, so as to provide sufficient data for subsequent model training and ensure the accuracy of model training.

[0119] Optionally, each set of second parameter information of each second switch in each second state obtained above, and the corresponding actual Ampere force and actual Holm force of this set of parameter information are stored as a data group, and then a data table can be obtained. Each row in the data table can be a data group, and different rows represent different second parameter information, actual Ampere force and actual Holm force.

[0120] Optionally, before S122 and S132, it further includes:

[0121] After obtaining multiple sets of second parameter information of each second switch in each second state, the data in each obtained data group is standardized to ensure data consistency and comparability. Standardization means converting data with different dimensions to the same dimension so that they have the same measurement standard. Each standardized data group is used to detect outliers by the interquartile range method, and the data group where the outlier is located is removed from the data table.

[0122] Clean the data groups in the data table to remove the error values and outliers. Error values refer to obvious errors in the data, such as negative conductivity values or values outside the reasonable range. In this embodiment, the outlier detection method uses statistical methods such as the interquartile range method to detect outliers L in the data. Specifically:

[0123]

[0124] ;

[0125] where Q1 is the first quartile, Q3 is the third quartile, and IQR is the interquartile range. is each second parameter information, actual Ampere force or actual Holm force after standardization processing.

[0126] S122: Use the current density of the second switch, the magnetic induction intensity of the second switch, the conductivity of the moving blade of the second switch, the electric field intensity of the second switch, the volume of the second switch, and the corresponding actual Ampere force as the first training data set.

[0127] The current density of the second switch, the magnetic induction intensity of the second switch, the conductivity of the moving blade of the second switch, the electric field intensity of the second switch, the volume of the second switch, and the corresponding actual Ampere force in each set of second parameter information of each second switch in each second state constitute a first training data set.

[0128] S132: Use the magnetic permeability of the moving blade of the second switch, the magnetic induction intensity of the second switch, the contact area between the moving blade and the static contact of the second switch, the density of the moving blade of the second switch, the current density of the second switch, the electric field intensity of the second switch, and the actual Holm force as the second training data set.

[0129] For each second switch, the permeability of the moving blade of the second switch, the magnetic induction intensity of the second switch, the contact area between the moving blade and the static contact of the second switch, the density of the moving blade of the second switch, the current density of the second switch, the electric field intensity of the second switch, and the actual Holm force in each set of second parameter information in each second state constitute a second training data set.

[0130] S142: Input c·m·a first training data sets into the first initial model to sequentially obtain c·m·a first initial Ampere forces; input c·m·a second training data sets into the second initial model to sequentially obtain c·m·a first initial Holm forces.

[0131] After each first training data set is input into the first initial model, the first initial model corresponds to an output value, and the output value of the first initial model is the first initial Ampere force. That is, when c·m·a first training data sets are input into the first initial model, c·m·a first initial Ampere forces are sequentially obtained. After each second training data set is input into the second initial model, the second initial model corresponds to an output value, and the output value of the second initial model is the first initial Holm force. That is, when c·m·a second training data sets are input into the second initial model, c·m·a first initial Holm forces are sequentially obtained.

[0132] S152: For each first initial Ampere force, after obtaining a first initial Ampere force, adjust the first weight according to the first initial Ampere force and the actual Ampere force corresponding to the first initial Ampere force; for each first initial Holm force, after obtaining a first initial Holm force, adjust the second weight according to the first initial Holm force and the actual Holm force corresponding to the first initial Holm force.

[0133] The actual Ampere force corresponding to the first initial Ampere force is the actual Ampere force in the first training data set corresponding to the first initial Ampere force, and the actual Holm force corresponding to the first initial Holm force is the actual Holm force in the second training data set corresponding to the first initial Holm force. The first electromotive force is the weighted sum of the first Ampere force and the first Holm force. That is, the sum of the product of the first Ampere force multiplied by the first weight and the product of the first Holm force multiplied by the second weight is determined as the first electromotive force of the first switch. Therefore, during the model training process, continuously adjust the magnitude of the first weight so that the first initial Ampere force is closer to the actual Ampere force, and adjust the magnitude of the second weight so that the first initial Holm force is closer to the magnitude of the actual Holm force.

[0134] According to the first initial Ampere force and the actual Ampere force, adjusting the first weight can be before adjusting the second weight according to the first initial Holm force and the actual Holm force, or can be after adjusting the second weight according to the first initial Holm force and the actual Holm force, and the sequence of the two is not limited.

[0135] S162: Obtain the first parameter information of the first switch in the first state.

[0136] S172: Input the current density of the first switch, the magnetic induction intensity of the first switch, the conductivity of the moving blade of the first switch, the electric field intensity of the first switch, and the volume of the first switch into the first target model to obtain the first Ampere force of the first switch.

[0137] S182: Input the magnetic permeability of the moving blade of the first switch, the magnetic induction intensity of the first switch, the contact area between the moving blade and the static contact of the first switch, the density of the moving blade of the first switch, the current density of the first switch, and the electric field intensity of the first switch into the second target model to obtain the first Holm force of the first switch.

[0138] S192: Determine the sum of the product of the first Ampere force multiplied by the first weight and the product of the first Holm force multiplied by the second weight as the first electrodynamic force of the first switch.

[0139] Optionally, adjusting the first weight according to the first initial Ampere force and the actual Ampere force includes:

[0140] 1) Use the ratio of the p-th actual Ampere force to the first initial Ampere force obtained by inputting the p-th first training data set into the first initial model as the first weight corresponding to the first initial model during the p-th training, and determine the value of the first loss function corresponding to the first weight during the p-th training; p is an integer greater than or equal to 1 and less than or equal to c·m·a.

[0141] Input the P-th training data set into the first initial model, and the output of the model is the first initial Ampere force obtained during the P-th training. Use the ratio of the p-th actual Ampere force to the first initial Ampere force obtained during the P-th training as the first weight corresponding to the first initial model during the p-th training.

[0142] Define a loss function during model training. The value of the loss function of the first initial model during the P-th training is the value of the first loss function. The specific calculation process of the value of the first loss function is described later.

[0143] 2) Input the second parameter information into the first policy network to obtain the first probability distribution, and the first probability distribution is used to represent the adjustment direction of the model parameters of the first initial model.

[0144] Figure 5This is a network architecture diagram provided by an embodiment of the present invention. Refer to Figure 5 In this stage, the model parameters will be dynamically optimized using a reinforcement learning network. Through continuous trial and feedback with the reinforcement learning algorithm, the model parameters are adjusted to minimize the prediction error and ensure that the model can maintain high accuracy under different conditions.

[0145] First, initialize the reinforcement learning environment. Take the calculation models of the Ampere force and the Holm force as the reinforcement learning environment, and define the state, action, and reward functions. Among them, the reinforcement learning environment is the place for the model parameters to interact and learn.

[0146] The state includes the model parameter settings of the first initial model and the prediction error:

[0147] ;

[0148] Among them, is the i-th model parameter, is the current prediction error. The current prediction error represents the deviation between the first initial Ampere force and the actual Ampere force. The state formula describes the current state of the first initial model, which is the input in reinforcement learning and provides a basis for subsequent action selection and policy optimization.

[0149] The action includes the magnitude and direction of adjusting the model parameters:

[0150] ;

[0151] Among them, is the adjustment amount for the i-th model parameter.

[0152] The reward function gives rewards or punishments according to the size of the prediction error. The smaller the error, the greater the reward:

[0153]

[0154] Among them, is the current prediction error. The smaller the error, the better the model performance and the greater the reward; is the weight parameter of the penalty term, which is used to balance the relationship between error reduction and the adjustment amplitude of the model parameters. If the value is too large, it may inhibit the adjustment effect; if the value is too small, it may cause excessive adjustment of the first initial model; is the adjustment amount for the i-th model parameter. Excessive adjustment may lead to model instability, so large parameter changes need to be punished. The reward function is the core of reinforcement learning and is used to evaluate whether a certain model parameter adjustment (action) is effective. The larger the reward value, the better the effect of the action.

[0155] Then, a complex multi-layer deep reinforcement learning network is used to optimize the parameters of the first policy network and the first value network through multi-task learning and adaptive gradient adjustment. The first policy network outputs an action vector for adjusting the model parameters:

[0156] ;

[0157] where s is the current state, is the first probability distribution output by the first policy network, is the weight of the first policy network, is the bias term of the first policy network, ReLU: activation function for introducing non-linearity, softmax: normalization function for converting network output into a probability distribution.

[0158] 3) Input the second parameter information into the first value network to obtain the first value, which is used to identify the adjustment value of the first policy network;

[0159] The first value network estimates the value of the current state to guide the optimization of the first policy network:

[0160]

[0161] where, is the first value output by the first value network, is the weight of the first value network, is the bias term of the first value network.

[0162] Policy update combines the first policy network and the first value network and is optimized through multi-task learning and adaptive gradient adjustment, where the loss function satisfies:

[0163] ;

[0164] In addition, an adaptive gradient adjustment mechanism is introduced to improve the optimization efficiency:

[0165] ;

[0166] ;

[0167] where, is the mean squared error of the first value network, is the weighted value network error of the logarithmic probability of the policy network, is the learning rate, is the sum of squared gradients of the policy network, is the cumulative sum of squared gradients of the value network, ϵ is a small constant to prevent division by zero, is the loss function. is the reward value, is the first value of the current state and the next state, is the discount factor, which weighs the relationship between the current reward and future rewards. The loss function consists of two parts:

[0168] The optimization objective of the first value network is to ensure accurate estimation of the first value;

[0169] : The optimization objective of the first policy network is to encourage the selection of actions that can maximize the value difference.

[0170] According to the formula of the above loss function, the corresponding first loss function value for each training is obtained. When the first initial model meets the training end condition, the first weight corresponding to the minimum first loss function value obtained from each training is used as the first weight of the first target model, and the corresponding first initial model is used as the first target model. Optionally, the training end condition can be that the number of training times of the first initial model reaches a preset value.

[0171] Optionally, according to the first initial Holm force and the actual Holm force, the second weight is adjusted, including:

[0172] 1) The ratio of the Pth actual Holm force to the first initial Holm force obtained by inputting the pth second training data set into the second initial model is used as the second weight corresponding to the second initial model during the Pth training, and the second loss function value corresponding to the second weight during the Pth training is determined;

[0173] 2) Input the second parameter information into the second policy network to obtain the second probability distribution, which is used to represent the adjustment direction of the model parameters of the second initial model;

[0174] 3) Input the second parameter information into the second value network to obtain the second value, which is used to identify the adjustment value of the second policy network.

[0175] The specific working processes of the above steps are similar to the specific process of adjusting the second weight according to the first initial Ampere force and the actual Ampere force, and will not be elaborated here.

[0176] This embodiment dynamically optimizes model parameters based on reinforcement learning, provides an intelligent adjustment mechanism for the prediction of electrodynamic force, and solves the problems of large prediction errors and weak model adaptability caused by fixed parameter settings in traditional methods. Through the reinforcement learning network, the model can perceive environmental changes in real time and dynamically adjust the weight of each parameter according to the input state (including model parameters and prediction errors), thereby reducing errors and improving accuracy. Under the conditions of complex electromagnetic field environment, vibration stress influence and variable material properties, the reinforcement learning mechanism allows the model to adjust the parameter configuration in an adaptive manner, realizes the global exploration of parameter optimization, and can quickly converge to the optimal solution, effectively avoiding the local optimum problem. In addition, the policy network and value network in this embodiment work together, guide the initial model to adjust parameters in the direction of reducing errors through the reward function, and combine the mechanism of adaptive learning rate to further improve the efficiency and stability of optimization. This dynamic learning characteristic enables the target model to remain efficient and accurate when facing complex and changeable scenarios in practical applications. Especially in the case of dynamic environmental changes or material performance degradation, it can quickly respond and make adjustments. Using reinforcement learning to train the model reduces the cumbersome manual parameter tuning, improves the optimization efficiency, unifies the optimization framework of multi-dimensional features, enables the solution to coordinately handle the non-linear relationship between multiple variables such as the electromagnetic field intensity of the switch, the electric field intensity of the switch, the current density of the switch, and the contact area between the moving blade and the static contact, and provides strong support for the entire technical solution. Through this intelligent and dynamic optimization method, not only significantly improves the prediction accuracy and robustness, but also expands the practical applicability of the target model in diverse scenarios, demonstrating extremely high technical innovation and application value.

[0177] An embodiment of the present invention further provides a device for determining electrodynamic force. Figure 6 The following is a schematic structural diagram of a device for determining electrodynamic force provided by an embodiment of the present invention. Refer to Figure 6 , the device for determining the electrodynamic force of the switch includes:

[0178] A parameter information acquisition module 1, configured to acquire first parameter information of a first switch in a first state; the first parameter information of the first switch in the first state includes: the magnetic induction intensity of the first switch, the electric field intensity of the first switch, the volume of the first switch, the current density of the first switch, the conductivity of the moving blade of the first switch, the magnetic permeability of the moving blade of the first switch, the density of the moving blade of the first switch, and the contact area between the moving blade and the static contact of the first switch.

[0179] A first Ampere force generation module 2, configured to input the current density of the first switch, the magnetic induction intensity of the first switch, the conductivity of the moving blade of the first switch, the electric field intensity of the first switch, and the volume of the first switch into a first target model to obtain the first Ampere force of the first switch.

[0180] The first Holm force generation module 3 is configured to input the magnetic permeability of the moving blade of the first switch, the magnetic induction intensity of the first switch, the contact area between the moving blade and the static contact of the first switch, the density of the moving blade of the first switch, the current density of the first switch, and the electric field intensity of the first switch into the second target model to obtain the first Holm force of the first switch.

[0181] The electrodynamic force generation module 4 is configured to determine the first electrodynamic force of the first switch according to the first Ampere force and the first Holm force.

[0182] The beneficial effects of the device for determining the electrodynamic force are the same as those of the method for determining the electrodynamic force of the switch, and will not be elaborated herein.

[0183] An embodiment of the present invention further provides an electronic device. Figure 7 The structural schematic diagram of the electronic device 10 that can be used to implement the embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as, laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as, personal digital processors, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are only examples and are not intended to limit the implementation of the present invention described and / or claimed herein.

[0184] As Figure 7 shown, the electronic device 10 includes at least one processor 11, and a memory communicatively connected to the at least one processor 11, such as a read-only memory (ROM), a random access memory (RAM), etc. The memory stores a computer program executable by the at least one processor. The processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) or the computer program loaded from the storage unit 18 into the random access memory (RAM). In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. The input / output (I / O) interface is also connected to the bus 14.

[0185] Multiple components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disc, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0186] The processor 11 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above.

[0187] An embodiment of the present invention also provides a computer-readable storage medium storing computer instructions for causing a processor to implement the method for determining the electrodynamic force of a switch in any of the above embodiments when executed.

[0188] The computer-readable storage medium can be a tangible medium that can contain or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. The computer-readable storage medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. Alternatively, the computer-readable storage medium can be a machine-readable signal medium. More specific examples of the machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0189] It should be understood that various forms of the processes shown above can be used, reordering, adding, or deleting steps. For example, the steps recited in the present invention can be executed in parallel, sequentially, or in a different order, as long as the desired results of the technical solution of the present invention can be achieved, and no limitation is made herein.

[0190] The above specific embodiments do not constitute a limitation on the protection scope of the present invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for determining the electrodynamic force of a switch, the switch including a moving blade and at least one stationary contact, characterized in that , The method for determining the electrodynamic force of the switch includes: Obtain the first parameter information of the first switch in the first state. The first parameter information of the first switch in the first state includes: the magnetic induction intensity of the first switch, the electric field intensity of the first switch, the volume of the first switch, the current density of the first switch, the conductivity of the moving blade of the first switch, the magnetic permeability of the moving blade of the first switch, the density of the moving blade of the first switch, and the contact area between the moving blade and the static contact of the first switch; Input the current density of the first switch, the magnetic induction intensity of the first switch, the conductivity of the moving blade of the first switch, the electric field intensity of the first switch, and the volume of the first switch into the first target model to obtain the first Ampere force of the first switch; Input the magnetic permeability of the moving blade of the first switch, the magnetic induction intensity of the first switch, the contact area between the moving blade and the static contact of the first switch, the density of the moving blade of the first switch, the current density of the first switch, and the electric field intensity of the first switch into the second target model to obtain the first Holm force of the first switch; Determine the first electrodynamic force of the first switch as the sum of the product of the first Ampere force multiplied by the first weight and the product of the first Holm force multiplied by the second weight.

2. The method for determining the electrodynamic force of the switch according to claim 1, characterized in that, Before obtaining the first parameter information of the first switch in the first state, it further includes: obtaining a set of second parameter information for each of the c second switches in each of the m second states, as well as the corresponding actual Ampere force and actual Holm force for each set of second parameter information; the second parameter information of the second switch in the second state includes: the magnetic induction intensity of the second switch, the electric field intensity of the second switch, the volume of the second switch, the current density of the second switch, the conductivity of the moving blade of the second switch, the magnetic permeability of the moving blade of the second switch, the density of the moving blade of the second switch, and the contact area between the moving blade and the static contact of the second switch; m, a, and c are all integers greater than or equal to 1; Use the current density of the second switch, the magnetic induction intensity of the second switch, the conductivity of the moving blade of the second switch, the electric field intensity of the second switch, the volume of the second switch, and the corresponding actual Ampere force as the first training data set; Use the magnetic permeability of the moving blade of the second switch, the magnetic induction intensity of the second switch, the contact area between the moving blade and the static contact of the second switch, the density of the moving blade of the second switch, the current density of the second switch, the electric field intensity of the second switch, and the actual Holm force as the second training data set; Input the c·m·a first training data sets into the first initial model to sequentially obtain c·m·a first initial Ampere forces; input the c·m·a second training data sets into the second initial model to sequentially obtain c·m·a first initial Holm forces; Adjust the first weight according to the first initial Ampere force and the actual Ampere force; adjust the second weight according to the first initial Holm force and the actual Holm force.

3. The method for determining the electrodynamic force of the switch according to claim 2, wherein adjusting the first weight according to the first initial Ampere force and the actual Ampere force includes: using the ratio of the p-th actual Ampere force to the first initial Ampere force obtained by inputting the p-th first training data set into the first initial model as the first weight corresponding to the first initial model during the p-th training, and determining the first loss function value corresponding to the first weight during the p-th training; p is an integer greater than or equal to 1 and less than or equal to c·m·a; inputting the second parameter information into the first policy network to obtain a first probability distribution, where the first probability distribution is used to represent the adjustment direction of the model parameters of the first initial model; inputting the second parameter information into the first value network to obtain a first value, where the first value is used to identify the adjustment value of the first policy network; adjusting the second weight according to the first initial Holm force and the actual Holm force includes: using the ratio of the p-th actual Holm force to the first initial Holm force obtained by inputting the p-th second training data set into the second initial model as the second weight corresponding to the second initial model during the p-th training, and determining the second loss function value corresponding to the second weight during the p-th training; inputting the second parameter information into the second policy network to obtain a second probability distribution, where the second probability distribution is used to represent the adjustment direction of the model parameters of the second initial model; inputting the second parameter information into the second value network to obtain a second value, where the second value is used to identify the adjustment value of the second policy network.

4. The method for determining the electrodynamic force of the switch according to any one of claims 1-3, characterized in that, The first parameter information of the first switch in the first state includes the electric field strength of the first switch; obtaining the first parameter information of the first switch in the first state includes: obtaining an electric field distribution image of the first switch in the first state; determining the electric field strength at the center position of the moving blade of the first switch according to the electric field distribution image, and using the electric field strength at the center position of the moving blade of the first switch as the electric field strength of the first switch.

5. The method for determining the electrodynamic force of the switch according to any one of claims 1-3, characterized in that, The first parameter information of the first switch in the first state includes the current density of the first switch; obtaining the first parameter information of the first switch in the first state includes: respectively obtaining the current values of the first switch at different times; generating an initial current waveform function according to the current values of the first switch at different times; generating a predicted current waveform function based on an ensemble learning algorithm and the initial current waveform function; determining the maximum value of the predicted current according to the predicted current waveform function, and determining the current density of the first switch according to the maximum value of the predicted current.

6. The method for determining the electrodynamic force of the switch according to any one of claims 1-3, characterized in that, The first parameter information of the first switch in the first state includes the density of the moving blade of the first switch, the conductivity of the moving blade of the first switch, and the magnetic permeability of the moving blade of the first switch; obtaining the first parameter information of the first switch in the first state includes: Determine the density, conductivity, and magnetic permeability of the moving blade of the first switch based on the material of the moving blade of the first switch and a preset corresponding relationship; wherein, the preset corresponding relationship is the corresponding relationship between the material and the density, conductivity, and magnetic permeability.

7. A device for determining the electrodynamic force of a switch, the switch comprising a moving blade and at least one stationary contact, characterized in that, The device for determining the electrodynamic force of the switch includes: A parameter information acquisition module, configured to acquire first parameter information of the first switch in a first state, where the first parameter information of the first switch in the first state includes: the magnetic induction intensity of the first switch, the electric field intensity of the first switch, the volume of the first switch, the current density of the first switch, the conductivity of the moving blade of the first switch, the magnetic permeability of the moving blade of the first switch, the density of the moving blade of the first switch, and the contact area between the moving blade and the static contact of the first switch; A first Ampere force generation module, configured to input the current density of the first switch, the magnetic induction intensity of the first switch, the conductivity of the moving blade of the first switch, the electric field intensity of the first switch, and the volume of the first switch into a first target model to obtain the first Ampere force of the first switch; A first Holm force generation module, configured to input the magnetic permeability of the moving blade of the first switch, the magnetic induction intensity of the first switch, the contact area between the moving blade and the static contact of the first switch, the density of the moving blade of the first switch, the current density of the first switch, and the electric field intensity of the first switch into a second target model to obtain the first Holm force of the first switch; An electrodynamic force generation module, configured to determine the sum of the product of the first Ampere force multiplied by a first weight and the product of the first Holm force multiplied by a second weight as the first electrodynamic force of the first switch.

8. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and when the computer program is executed by the at least one processor, the at least one processor is enabled to execute the method for determining the electrodynamic force of the switch according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for enabling a processor to implement the method for determining the electrodynamic force of the switch according to any one of claims 1-6 when executed.

Images