A multi-piecewise linear modeling solution method for switching transients in power electronic converters

Through the multi-section linear modeling method, the transient process of SiC MOSFET switch is divided into three stages and non-iterative solutions are performed. The problems of simulation accuracy and calculation complexity in the existing technology are solved, and the SiC MOSFET switch process is realized. efficient, accurate description and optimization of power electronic converter.

CN118627450BActive Publication Date: 2024-11-22NORTHWESTERN POLYTECHNICAL UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411109488.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-13
Publication Date
2024-11-22
Estimated Expiration
2044-08-13

AI Technical Summary

Technical Problem

When evaluating and optimizing the switch transient process of SiC MOSFET, the prior art has problems such as insufficient simulation accuracy, large calculation amount and complex calculations, making it difficult to quickly and accurately describe and predict switch losses and non-ideal behaviors.

Method used

The multi-segment linear modeling method is used to divide the switch transient process into three stages, and an equivalent circuit model is established through non-iteration solutions, and the numerical integral solution is used to solve the problem by backward Euler's method, which simplifies the calculation process.

Benefits of technology

It realizes a high-accuracy description of the transient process of SiC MOSFET switches, reduces calculation complexity and time, improves simulation efficiency, and is suitable for the design and optimization of power electronic converters.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118627450B_ABST
    Figure CN118627450B_ABST
Patent Text Reader

Abstract

The present invention proposes a multiple piecewise linear modeling and solving method and system for the switching transient of a power electronic converter, using a half-bridge converter as a calculation unit. The method first constructs a behavioral model according to the transient process of the switch tube module, and then divides the switching transient into three different stages according to the state of the switch tube and the diode, and establishes an equivalent circuit for each stage. The equivalent circuit of each stage is solved non-iteratively by a multiple piecewise linear method. The present invention successfully applies multiple piecewise linear models to solve circuits in buck converters, demonstrating its wide application potential in the design and optimization of power electronic converters. Compared with traditional methods, the present invention can reduce the computational complexity while performing transient simulation more accurately, achieving efficient and accurate simulation analysis, and helping to optimize the design and improve system performance.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of power electronic simulation, and in particular relates to a method for solving transient multi-piecewise linear modeling of a power electronic converter switch. Background Art

[0002] In recent years, with the growing application demand in new energy and other industries, a new generation of wide bandgap power semiconductor devices represented by SiC MOSFET has developed rapidly. Its high switching speed has brought new development opportunities for the high frequency and high power density of power electronics technology. However, when it is applied to power electronic converters, how to evaluate and optimize the switching transient process of SiCMOSFET, which only lasts for a dozen nanoseconds or even a few nanoseconds, including predicting and limiting switching losses and non-ideal behaviors including oscillations, transient voltage / current overshoot spikes, etc., has become a huge challenge. In order to evaluate power electronic converters to improve the application efficiency of silicon carbide switching tubes and reduce costs, it is urgent to establish a high-precision, low-computational-overhead SiC MOSFET simulation model.

[0003] At present, the modeling methods of power semiconductors can be classified into two categories according to the different application levels of the models: system-level models and device-level models. System-level models include average models, switch function models, ideal switch models, binary resistance models, constant admittance L / C models, etc. This type of model only focuses on the impact of the switch steady state on the circuit, and regards the power semiconductor as two states of the switch without paying attention to the characteristics of the switch switching process. The device-level model includes two categories: physical models and behavioral models. The physical model describes the carrier distribution and its electrical characteristics based on the semiconductor physics equation. The physical model usually involves partial differential equations or nonlinear differential equations, so it has a high computational complexity, and the parameters involve the device manufacturing process, and the extraction process is very complicated. The behavioral model does not involve complex physical mechanisms. It uses macroscopic models and empirical formulas to describe the switching characteristics and the parameters are easier to extract. When the device-level model is applied to power electronic converters, the switching device is no longer regarded as an ideal model and its dynamic characteristics are ignored. On the one hand, the transient waveform of the switching device can be provided to realize the state monitoring of the switching device during the operation of the power electronic converter and realize the health management of the switching device; on the other hand, the influence of the dynamic characteristics such as electrical parameter oscillation and spikes of the switching device during the transient process on the power electronic converter can be characterized to achieve more accurate control of the converter. In order to accurately describe the switching transient process of SiC MOSFET while reducing complexity, it is essential to use a behavioral model and simplify it.

[0004] In the research on analytical modeling of SiC MOSFET switching transient, the simplified methods introduced can be summarized into the following three types: from the perspective of simplifying the switching transient process, by assuming the segmentation and linearity of the waveform, the physical mechanism is effectively decoupled, so that the complex physical process can be simplified; from the perspective of the circuit, some stray parameters that have little effect on the switching transient process are ignored to simplify the circuit structure; from the perspective of parameters, the modeling of key parameters such as parasitic capacitance and transconductance is simplified, and then the equations are simplified. At present, there are numerical models, segmented models and traditional piecewise linear models for simplifying the SiC MOSFET behavior model. In order to retain the accuracy of the model to the maximum extent, the numerical model uses the expressions of channel current and parasitic capacitance in the physical model or fits the relevant curves in the device data sheet, and rewrites the original differential equations into the form of state equations, and solves them iteratively. The final model relies on the data processing capabilities of software such as Matlab. Although it can accurately model the switching transient process, it lacks intuitiveness. In the piecewise model, the function form of each component is clear and unambiguous. It can better reflect the influence of each system element on the switching transient while ensuring high accuracy. It is suitable for high-accuracy evaluation of switching loss and precise evaluation of non-ideal behavior. However, solving the differential equation still requires more calculations. The resulting voltage and current expressions contain exponential functions, trigonometric functions, and multiple components, which are still relatively complex and not suitable for applications such as rapid loss evaluation. The traditional piecewise linear model is extremely practical, concise in form, and easy to establish. It is widely used in rapid evaluation of switching loss. However, due to the neglect of many characteristics of the actual waveform, the model deviates greatly from the actual waveform and the accuracy is not ideal.

[0005] It can be seen that the current switching transient modeling and solving methods have the problems of insufficient simulation accuracy, large amount of calculation, and complex calculation. Using appropriate methods to achieve fast and accurate solution of switching transients is crucial to the development of power electronics technology. Summary of the invention

[0006] The problem to be solved by the present invention is to provide a method for solving the switching transient of a power electronic converter by multiple piecewise linear modeling in view of the deficiencies of the above-mentioned prior art, and specifically relates to a method for solving the switching transient modeling of a power electronic converter by multiple piecewise linearity with low complexity and high accuracy. From the perspective of the switching transient process, a staged solution method is adopted to divide the switching transient into three stages for solution; from the perspective of the circuit, a nonlinear characteristic piecewise linear method is adopted to perform piecewise linearization on the nonlinear characteristics of the simulated current source of the simulated transfer characteristics; from the perspective of parameters, a method for piecewise constantization of time-varying parameters is adopted to perform piecewise linearization on the variable capacitor. The specific steps are as follows:

[0007] Step A1, determining a calculation unit, wherein the switch transient modeling calculation unit is a half-bridge converter;

[0008] Step A2: Build a behavior model based on the switching transient process, and build a behavior model that can accurately represent the switching transient according to the switching transient waveform and physical effects of the switch tube and its body diode;

[0009] Step A3: Divide the switching transient into three different stages according to the state of the switch tube and the diode and establish equivalent circuits for each of them; wherein the switching transient selects the diode reverse process in each stage to simulate the inductor current , drain current , gate voltage , drain-source voltage Four variables are used as state variables, and the three stages specifically include:

[0010] Stage 1: The diode is forward conducting. The ideal diode in the diode model is equivalent to a voltage source and a resistor in series, where the voltage value of the voltage source is the forward voltage drop of the diode. The resistance value is the forward resistance of the diode , write the current equation for the diode model, there is a relationship:

[0011] ,

[0012] The input voltage source for this stage The Kirchhoff voltage equation can be written as follows for the main circuit consisting of a diode and a switch:

[0013] ,

[0014] in, is the drain-source voltage in the steady state when the switch is turned off. Under this equivalent condition, its expression is ; Indicates the input current, represents the equivalent inductance of stray inductance, represents the inductor simulating the reverse recovery characteristics, Represents the resistance simulating the reverse recovery characteristic;

[0015] The gate voltage of the switch tube is lower than the threshold voltage , in the cut-off region, the analog current source used to simulate the dynamic characteristics of the switch tube At this stage, writing the current equation for the gate can be obtained:

[0016] ,

[0017] in, is the gate control voltage, is the gate resistance; is the equivalent variable capacitance of the parasitic capacitance between drain and gate, is the equivalent capacitance of the parasitic capacitance between the gate and the source;

[0018] Writing the current equation for the source gives:

[0019] ,

[0020] in, The equivalent variable capacitance represents the parasitic capacitance between drain and source;

[0021] Phase 2: The diode is forward-conducting, the gate voltage of the switch is higher than the threshold voltage, and the analog current source With gate voltage and drain-source voltage Related, in accordance with the functional relationship: , write the current equation for the diode model at this stage, and there is a relationship:

[0022] ,

[0023] Writing Kirchhoff's voltage equation for the main circuit of this stage gives:

[0024] ,

[0025] Writing the current equation for the gate gives:

[0026] ,

[0027] Writing the current equation for the source gives:

[0028] ;

[0029] Stage 3: The diode is reversely turned off, and the ideal diode in the diode model is equivalent to an open circuit; the gate voltage of the switch tube is higher than the threshold voltage, and the simulated current source is related to the gate voltage and the drain-source voltage, which conforms to the functional relationship: ;

[0030] At this stage, since the ideal diode is equivalent to an open circuit, the reverse recovery simulation resistance and simulation inductance in the diode model conform to the relationship:

[0031] ,

[0032] Diode module current Conforms to the relationship:

[0033] ,

[0034] in, represents a controlled current source simulating reverse recovery characteristics, K is the controlled current source coefficient;

[0035] The relationship between the diode current and the drain current is as follows:

[0036] ,

[0037] Writing the current equation for the gate in the equivalent circuit of the switch tube gives:

[0038] ,

[0039] Writing the current equation for the source gives:

[0040] ;

[0041] Step A4, performing non-iterative solutions to the equivalent circuits of each stage;

[0042] Step A5: Apply the multiple piecewise linear models to solve the circuit of the buck converter.

[0043] Furthermore, in step A1, the half-bridge converter includes two upper and lower SiC MOSFET / diode pairs. In the switching event, the upper and lower SiC MOSFETs are not turned on at the same time. It can be simplified to that the upper switch tube and the lower diode are always in the off state, which is equivalent to an open circuit. Only the upper diode and the lower switch tube are affected during the switching process. And in the transient calculation process, the input voltage of the calculation unit is calculated. and input current Considered as a constant value.

[0044] Furthermore, the behavioral model constructed in step A2 can accurately represent the output characteristics and transfer characteristics of the cut-off region and the non-cut-off region of the switching tube in the transient process with low complexity, and can accurately represent the reverse recovery characteristics of the diode. The model of the switching tube SiCMOSFET consists of the equivalent inductance of the stray inductance , the equivalent variable capacitance of the parasitic capacitance between drain and source , the equivalent variable capacitance of the parasitic capacitance between drain and gate , the equivalent capacitance of the parasitic capacitance between the gate and source And a controlled current source for simulating the dynamic characteristics of the switch tube The diode model consists of an ideal diode and an inductor that simulates the reverse recovery characteristics. ,resistance , controlled current source composition.

[0045] Furthermore, the specific steps of performing non-iterative solutions to the equivalent circuits in step A4 include:

[0046] Step B1, loading the calculation result of the previous step;

[0047] Step B2: Determine which stage is in the process of opening, the gate control voltage , the switch tube switches from the cut-off state to the on state, and the diode switches from the on state to the cut-off state. First, determine whether the switch tube is in the cut-off region. The judgment basis is whether the gate voltage reaches the threshold voltage: If , then it is in stage 1, the analog current source , and then judge the diode state. The opening process adopts the current judgment method: if the current flowing through the diode model is greater than the controlled current source current, i.e. , the ideal diode in the diode model remains on and is in stage two; otherwise, the ideal diode is disconnected and is in stage three. During the shutdown process, the gate control voltage , the switch tube switches from the on state to the off state, and the diode switches from the off state to the on state. First, the diode state is judged, and the shutdown process adopts the voltage judgment method: if , the ideal diode in the diode model withstands the reverse voltage and remains disconnected, which is in stage three; otherwise, the ideal diode is turned on, and then the switch state is judged. If , the switch tube has not yet reached the cut-off zone and is in stage 2; otherwise, it switches to the cut-off state and is in stage 1;

[0048] Step B3: Calculate and update the value of equivalent parasitic capacitance. is a constant value, and They are all variable capacitors, and their values ​​change with the voltage across the capacitor, which conforms to the functional relationship , ,in ;

[0049] Step B4: Use a lookup table to determine the value of the analog current source. The specific implementation method is to first linearize the analog current source and convert the original complex nonlinear function Expressed as of the form, where , and is the linearization coefficient. According to the principle of multivariate function linearization, When the linearization coefficient For the original complex nonlinear function This point is The partial guide, For The partial guide, is a constant term, and its expression is , , , then select the appropriate and Segment distance, the function Divide into multiple planes, and store the linearization results in the form of a two-dimensional table. The value of is used as the index to determine , and The value of

[0050] Step B5: Write the correct state equation according to different stages to solve the circuit and get the matrix form of the state equation. ,in, is the state variable vector, is the input vector, the coefficient matrix is the state matrix, the coefficient matrix To control the matrix, first define the coefficients , , , , , , for stage 1, the matrix form of the state equation can be expressed as:

[0051] ,

[0052] Among them, the diode reverse process simulates the inductor current , drain current , gate voltage , drain-source voltage Four variables are state variables;

[0053] The expression of the coefficient matrix is:

[0054] State Matrix ,

[0055] Control Matrix ;

[0056] For stage 2, the matrix form of the state equation can be expressed as:

[0057] ,

[0058] The expression of the coefficient matrix is:

[0059] State Matrix ,

[0060] Control Matrix ,

[0061] Where K is the controlled current source coefficient, is the linearization coefficient of the current source in the equivalent circuit of the switch tube;

[0062] For stage three, the diode state equation is:

[0063] ;

[0064] The calculation method of the switch tube drain current is:

[0065] ,

[0066] The matrix form of the switch state equation can be expressed as:

[0067] ,

[0068] in, , ;

[0069] The state equations obtained for different stages are The numerical integration is solved using the backward Euler method, and the time step used in the calculation process is , then the state variable value of the current step length is expressed as Update the calculation results, where: is the identity matrix, is the state variable vector of the current step, is the state variable vector of the next step length;

[0070] Step B6: Update the network with the solution result, end the current step length, and wait for the next step length to start.

[0071] Furthermore, in step A5, the multiple piecewise linear model is applied to the circuit solution of the buck converter. The device-level simulation of the power electronic converter in the buck converter means solving the switching devices in the buck converter using the multiple piecewise linear method, and solving other linear parts of the circuit according to the backward Euler method.

[0072] The present invention also provides a power electronic converter switch transient multi-piecewise linear modeling system, the system comprising:

[0073] A calculation unit module, used for determining to use a half-bridge converter as a calculation unit;

[0074] Behavior model building module, used to build a behavior model according to the transient process of the switch tube module;

[0075] An equivalent circuit establishment module is used to divide the switching transient state into three different stages according to the states of the switch tube and the diode and establish equivalent circuits for each of them;

[0076] A non-iterative solution module is used to perform non-iterative solutions to the equivalent circuits of each stage by using a multiple piecewise linear method;

[0077] Application module for applying multiple piecewise linear models to circuit solving of buck converters.

[0078] Furthermore, the computing unit module includes:

[0079] The half-bridge converter includes two upper and lower SiC MOSFET / diode pairs. The upper switch tube and the lower diode are simplified to be in the off state, which is equivalent to an open circuit. Only the upper diode and the lower switch tube are affected during the switching process, and the input voltage and input current of the calculation unit are regarded as constants during the transient calculation process.

[0080] Furthermore, the behavioral model building module can accurately represent the output characteristics and transfer characteristics of the cut-off region and the non-cut-off region during the transient process of the switching tube, as well as the reverse recovery characteristics of the diode with low complexity. The model of the switching tube SiC MOSFET is composed of the equivalent inductance of the stray inductance, the equivalent variable capacitance of the parasitic capacitance between the drain and the source, the equivalent variable capacitance of the parasitic capacitance between the drain and the gate, the equivalent capacitance of the parasitic capacitance between the gate and the source, and a controlled current source for simulating the dynamic characteristics of the switching tube. The diode model is composed of an ideal diode and an inductor, a resistor, and a controlled current source for simulating the reverse recovery characteristics.

[0081] Furthermore, the equivalent circuit establishment module divides the switching transient into three stages, selects four variables of the diode reverse process simulation inductor current, drain current, gate voltage, and drain-source voltage as state variables in each stage, and establishes equivalent circuits of stage one, stage two, and stage three respectively, writes the correct state equation solving circuit for different stages, and obtains the matrix form of the state equation ,in, is the state variable vector, is the input vector, the coefficient matrix is the state matrix, the coefficient matrix The control matrix is ​​solved by numerical integration using the backward Euler method. Using the backward Euler method, the state variable value of the current step size is Update the calculation results and use the solution results to update the network, end the current step, and wait for the next step to start.

[0082] Compared with the prior art, the present invention has the following advantages:

[0083] 1. The model design of the method of the present invention conforms to the output characteristics of actual power semiconductor devices, and can accurately describe the transient characteristics that can only be described by non-ideal models such as the Miller platform of SiCMOSFET and the reverse recovery of the diode, with high accuracy.

[0084] Compared with the iterative solution simulation method of the existing power electronic design software, the method of the present invention realizes non-iterative solution by using appropriate segmentation and linearization in the process of circuit equivalent simplification and parameter equivalent substitution, which greatly reduces the calculation time and solves the calculation cost.

[0085] When the model constructed by the method of the present invention is simulated under different working conditions, the model only needs to change the input parameters, and has strong portability; for other power switching devices with transient characteristics or physical parameter changes, building multiple piecewise linear models only requires modifying some parameters without changing the overall model, and has strong versatility.

[0086] The methods such as multivariate function linearization and data pre-storage used in the present invention provide an effective solution for the rapid transient solution of power semiconductors.

[0087] The method of the present invention can be quickly applied to the device-level simulation of the power electronic converter, and provides the transient characteristics of the switching device during the simulation of the power electronic converter, which is helpful for analyzing the loss and health status of the switching device.

[0088] In summary, the method for solving transient modeling of multiple piecewise linear power semiconductor switches provided by the present invention has the advantages of high accuracy, short calculation time, strong portability, strong versatility, etc. The method is reasonably designed and simple to implement.

[0089] 2. This invention proposes a multi-piecewise linear modeling solution method for the problem of switching transient modeling in power electronic converters. This method establishes an accurate behavioral model by analyzing the behavioral characteristics of the switch tube and the diode in the transient process in detail, and constructs an equivalent circuit and performs non-iterative solution in stages. This technology solves the shortcomings of traditional modeling methods in terms of accuracy and efficiency, and provides a more accurate and efficient tool for the design and optimization of power electronic converters.

[0090] In industrial applications, this modeling and solving method significantly improves the design quality and efficiency of power electronic converters. Traditional simulation methods often fail to accurately capture the details of the switching transient process, resulting in a large difference between the actual operation results and the simulation predictions. The method proposed in this invention can simulate the dynamic behavior of the switching device more realistically, helping engineers to accurately predict the performance of the converter in the design stage, thereby effectively reducing the workload of later debugging and optimization.

[0091] In addition, this method has good versatility and scalability, and can be widely used in different types of power electronic converters, such as buck converters, boost converters, etc. By using this method to perform device-level simulation, the performance of the converter under different working conditions can be more accurately evaluated, providing strong support for product development and improvement.

[0092] In summary, the present invention has made significant technological progress in the field of power electronic converters, which not only improves the accuracy and efficiency of simulation, but also brings convenience to the practical application of the industry. Through this modeling and solving method, enterprises can more quickly develop power electronic converter products with superior performance, stability and reliability to meet the ever-changing market needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0093] Figure 1 A flow chart of the method for solving the transient modeling of a switch according to the present invention;

[0094] Figure 2 This is a schematic diagram of the stage switching conditions of the present invention;

[0095] Figure 3 It is the equivalent circuit diagram of three stages of the present invention;

[0096] Figure 4 This is a comparison diagram of gate voltage simulation results of the present invention;

[0097] Figure 5 A comparison diagram of gate voltage simulation results in the conduction transient state of the present invention;

[0098] Figure 6 A comparison diagram of gate voltage simulation results in the turn-off transient state of the present invention;

[0099] Figure 7 This is a comparison diagram of the drain-source voltage simulation results of the present invention;

[0100] Figure 8 This is a comparison diagram of transient simulation results of drain-source voltage in the conduction state of the present invention;

[0101] Fig. 9 This is a comparison diagram of transient simulation results of the transient drain-source voltage when the present invention is turned off;

[0102] Fig.10 This is a comparison diagram of transient simulation results of drain current of the present invention;

[0103] Fig.11 This is a comparison diagram of transient simulation results of the transient drain current in the conduction state of the present invention;

[0104] Fig.12 This is a comparison diagram of transient simulation results of the transient drain current when the present invention is turned off;

[0105] Fig.13 This is a model diagram of a buck converter of the present invention;

[0106] Fig.14 This is a comparison diagram of the output voltage simulation results of the buck converter of the present invention;

[0107] Fig.15 It is a comparison diagram of the inductor current simulation results of the buck converter of the present invention. DETAILED DESCRIPTION

[0108] The following are two industrial application examples, which demonstrate the application and effects of the present invention in actual industrial scenarios:

[0109] Example 1: Electric vehicle charging station

[0110] In electric vehicle charging stations, power electronic converters play a vital role, especially in fast charging technology. By using the switch transient modeling solution method of the present invention, the design and operation efficiency of the charging station can be significantly improved.

[0111] 1. Accurate simulation and design optimization:

[0112] Using the modeling method of the present invention, engineers can accurately simulate the switching transient behavior of the power electronic converter used in the charging station.

[0113] This helps to accurately predict and optimize charging station performance during the design phase, especially under high-power, high-efficiency charging demands.

[0114] 2. Reduce debugging time and cost:

[0115] By using the modeling and solving method of the present invention in the simulation stage, potential design problems can be discovered and solved before actual deployment.

[0116] This greatly reduces the complexity and time cost of on-site debugging and improves the construction efficiency of charging stations.

[0117] 3. Improve system stability and reliability:

[0118] Accurate switching transient models help identify potential failure modes and system weaknesses.

[0119] By considering these factors during the design phase, a more stable and reliable charging system can be built, reducing failures and maintenance requirements during operation.

[0120] Embodiment 2: Wind power generation system

[0121] In a wind power generation system, a power electronic converter is used to convert the variable frequency electric energy generated by a wind power generator into a stable grid frequency electric energy. The switch transient modeling and solving method of the present invention also has significant application value here.

[0122] 1. High-efficiency conversion design:

[0123] The modeling method of the present invention can accurately simulate the dynamic behavior of the power electronic converter in the wind power generation system.

[0124] This helps design more efficient and stable power conversion schemes, thereby improving the efficiency of the entire wind power generation system.

[0125] 2. Reduce energy consumption and operating costs:

[0126] By optimizing the design of power electronic converters, energy consumption and heat generation during system operation can be reduced.

[0127] This not only reduces operating costs, but also helps improve the environmental performance of the system.

[0128] 3. Enhance system robustness:

[0129] Accurate switching transient models can reveal the system behavior under different wind and grid conditions.

[0130] This enables engineers to design more robust systems to cope with a variety of complex and uncertain operating environments.

[0131] In summary, the switch transient modeling and solving method of the present invention has broad application prospects in industrial fields such as electric vehicle charging stations and wind power generation systems, and can significantly improve the design efficiency, stability and operation efficiency of the system.

[0132] The present invention relates to a multi-piecewise linear modeling system for power electronic converter switch transient and a solution method thereof, which is mainly used to solve the problems of complex modeling and low computational efficiency of power electronic converter switch transient in the prior art. By dividing the switch transient process into three different stages and establishing an equivalent circuit model for each stage, accurate simulation of the switch transient is achieved. The system can not only improve the accuracy of the model, but also effectively reduce the computational complexity and improve the computational efficiency.

[0133] In the prior art, the modeling of switching devices often requires solving complex nonlinear equations, and the problem of slow or even non-convergence of iterations is prone to occur during the calculation process. The present invention adopts a multiple piecewise linear method to divide the switching transient process into multiple linear sections, and each section is solved non-iteratively, thereby avoiding the problems that occur during the iteration process.

[0134] The significant technical progress of the present invention is that the proposed multiple piecewise linear modeling method can accurately represent the transient process of the switch tube with low complexity, including the output characteristics and transfer characteristics of the cut-off region and the non-cut-off region, and the reverse recovery characteristics of the diode. By modeling the states of the switch tube and the diode separately and combining the backward Euler method for numerical integration and solution, the present invention realizes high-precision simulation of the switching transient process and can quickly obtain the solution result, thereby improving the simulation efficiency and accuracy in industrial applications.

[0135] In industrial applications, the multi-segmented linear modeling system and solution method of the present invention can be widely used in the design and optimization of power electronic converters. In the buck converter, the method of the present invention can replace the traditional modeling and solution method, and perform accurate transient simulation of the switching device in the buck converter while maintaining the linear solution of other parts of the circuit, thereby achieving efficient and accurate simulation analysis in practical engineering applications. Through the method of the present invention, engineers can more accurately predict the performance of switching devices, optimize the design of power electronic converters, and improve the overall performance and reliability of the system.

[0136] like Figure 1 As shown, the switch transient modeling and solving method of the present invention includes the following steps:

[0137] A1. Determine the calculation unit and use the half-bridge converter as the calculation unit;

[0138] A2. Build a behavioral model based on the transient process of the switch module;

[0139] A3. Divide the switching transient state into three different stages according to the state of the switch tube and the diode and establish equivalent circuits for each of them;

[0140] A4. Use multiple piecewise linear methods to solve the equivalent circuits of each stage non-iteratively.

[0141] A5. Apply multiple piecewise linear models to solve the circuit of the buck converter.

[0142] In this embodiment, the half-bridge converter in step A1 includes two upper and lower SiC MOSFET / diode pairs. In the switching event, the upper and lower SiC MOSFETs are not turned on at the same time. It can be simplified as the upper switch tube and the lower diode are always in the off state, which is equivalent to an open circuit. Only the upper diode and the lower switch tube are affected in the switching process. And in the transient calculation process, the input voltage of the calculation unit is calculated. and input current Considered as a constant value.

[0143] In this embodiment, the behavioral model constructed in step A2 can accurately represent the output characteristics and transfer characteristics of the cut-off region and the non-cut-off region of the switching tube in the transient process with low complexity, and can accurately represent the reverse recovery characteristics of the diode. The model of the switching tube SiC MOSFET is composed of the equivalent inductance of the stray inductance , the equivalent variable capacitance of the parasitic capacitance between drain and source , the equivalent variable capacitance of the parasitic capacitance between drain and gate , the equivalent capacitance of the parasitic capacitance between the gate and source And a controlled current source for simulating the dynamic characteristics of the switch tube The diode model consists of an ideal diode and an inductor that simulates the reverse recovery characteristics. ,resistance , controlled current source composition.

[0144] In this embodiment, Figure 2 As shown, the switching transient in step A3 is divided into three stages, and the reverse process of the diode is selected in each stage to simulate the inductor current , drain current , gate voltage , drain-source voltage Four variables are used as state variables, and the three stages specifically include:

[0145] Stage 1: The diode is forward conducting. The ideal diode in the diode model is equivalent to a voltage source and a resistor in series, where the voltage value of the voltage source is the forward voltage drop of the diode. The resistance value is the forward resistance of the diode , write the current equation for the diode model, there is a relationship:

[0146] ,

[0147] The input voltage source for this stage The Kirchhoff voltage equation can be written as follows for the main circuit consisting of a diode and a switch:

[0148] ,

[0149] in, is the drain-source voltage in the steady state when the switch is turned off. Under this equivalent condition, its expression is ;

[0150] The gate voltage of the switch tube is lower than the threshold voltage , in the cut-off region, the analog current source used to simulate the dynamic characteristics of the switch tube At this stage, writing the current equation for the gate can be obtained:

[0151] ,

[0152] in, is the gate control voltage, is the gate resistance;

[0153] Writing the current equation for the source gives:

[0154] ;

[0155] Phase 2: The diode is forward-conducting, the gate voltage of the switch is higher than the threshold voltage, and the analog current source With gate voltage and drain-source voltage Related, in accordance with the functional relationship: , write the current equation for the diode model at this stage, and there is a relationship:

[0156] ,

[0157] Writing Kirchhoff's voltage equation for the main circuit of this stage gives:

[0158] ,

[0159] Writing the current equation for the gate gives:

[0160] ,

[0161] Writing the current equation for the source gives:

[0162] ;

[0163] Stage 3: The diode is reversely turned off, and the ideal diode in the diode model is equivalent to an open circuit; the gate voltage of the switch tube is higher than the threshold voltage, and the simulated current source is related to the gate voltage and the drain-source voltage, which conforms to the functional relationship: ;

[0164] At this stage, since the ideal diode is equivalent to an open circuit, the reverse recovery simulation resistance and simulation inductance in the diode model conform to the relationship:

[0165] ,

[0166] Diode module current Conforms to the relationship:

[0167] ,

[0168] The relationship between the diode current and the drain current is as follows:

[0169] ,

[0170] Writing the current equation for the gate in the equivalent circuit of the switch tube gives:

[0171] ,

[0172] Writing the current equation for the source gives:

[0173] .

[0174] In this embodiment, Figure 3As shown, the specific steps of solving step A4 include:

[0175] Step B1, loading the calculation result of the previous step;

[0176] Step B2: Determine which stage is in the process of opening, the gate control voltage , the switch tube switches from the cut-off state to the on state, and the diode switches from the on state to the cut-off state. First, determine whether the switch tube is in the cut-off region. The judgment basis is whether the gate voltage reaches the threshold voltage: If , then it is in stage 1, the analog current source , and then judge the diode state. The opening process adopts the current judgment method: if the current flowing through the diode model is greater than the controlled current source current, i.e. , the ideal diode in the diode model remains on and is in stage two; otherwise, the ideal diode is disconnected and is in stage three. During the shutdown process, the gate control voltage , the switch tube switches from the on state to the off state, and the diode switches from the off state to the on state. First, the diode state is judged, and the shutdown process adopts the voltage judgment method: if , the ideal diode in the diode model withstands the reverse voltage and remains disconnected, which is in stage three; otherwise, the ideal diode is turned on, and then the switch state is judged. If , the switch tube has not reached the cut-off region and is in stage two; otherwise, it switches to the cut-off state and is in stage one.

[0177] The judgment method is shown in the following table:

[0178] Table 1 Basis for switching at different stages

[0179] Phase 1 Phase 2 Phase 3 Ideal diode state Conductivity Conductivity Shutdown MOSFET Status Cut-off area Linear region, saturation region Linear region, saturation region When Vg>0, the end mark Vgs>Vth Id<KVlrr - When Vg≤0, the start flag - Vgs>Vth Vds≤Vcc+Vd_on

[0180] Step B3: Calculate and update the value of equivalent parasitic capacitance. is a constant value, and They are all variable capacitors, and their values ​​change with the voltage across the capacitor, which conforms to the functional relationship , ,in ;

[0181] Step B4: Use a lookup table to determine the value of the analog current source. The specific implementation method is to first linearize the analog current source and convert the original complex nonlinear function Expressed as of the form, where , and is the linearization coefficient. According to the principle of multivariate function linearization, When the linearization coefficient For the original complex nonlinear function This point is The partial guide, For The partial guide, is a constant term, and its expression is , , , then select the appropriate and Segment distance, the function Divide into multiple planes, and store the linearization results in the form of a two-dimensional table. The value of is used as the index to determine , and The value of

[0182] Step B5: Write the correct state equation according to different stages to solve the circuit and get the matrix form of the state equation. ,in, is the state variable vector, is the input vector, the coefficient matrix is the state matrix, the coefficient matrix is the control matrix; for the convenience of expression, first define the coefficient , , , , , , for stage 1, the matrix form of the state equation can be expressed as:

[0183] ,

[0184] Among them, the diode reverse process simulates the inductor current , drain current , gate voltage , drain-source voltage Four variables are state variables;

[0185] The expression of the coefficient matrix is:

[0186] State Matrix ,

[0187] Control Matrix ;

[0188] For stage 2, the matrix form of the state equation can be expressed as:

[0189] ,

[0190] The expression of the coefficient matrix is:

[0191] State Matrix ,

[0192] Control Matrix ,

[0193] Where K is the controlled current source coefficient, is the linearization coefficient of the current source in the equivalent circuit of the switch tube;

[0194] For stage three, the diode state equation is:

[0195] ;

[0196] The calculation method of the switch tube drain current is:

[0197] ,

[0198] The matrix form of the switch state equation can be expressed as:

[0199] ,

[0200] in, , ;

[0201] The state equations obtained for different stages are The numerical integration is solved using the backward Euler method, and the time step used in the calculation process is , then the state variable value of the current step length is expressed as Update the calculation results, where: is the identity matrix, is the state variable vector of the current step, is the state variable vector of the next step length;

[0202] Step B6: Update the network with the solution result, end the current step length, and wait for the next step length to start.

[0203] In this real-time example, the above-mentioned switching transient model solving method is applied to the solution of the power electronic converter. The power electronic converter device-level simulation of the buck converter in step A5 means replacing the switching device in the buck converter with the above-mentioned modeling and solving method, and the other linear parts of the circuit are still solved according to the backward Euler method.

[0204] In order to verify the effectiveness of the modeling and solving method proposed in this invention, a simulation comparison was carried out.

[0205] Taking the C3M0060065K module as an example, a behavioral model is established and non-iterative solution is performed. The reference model is a simplified version of the LTspice simulation model officially provided by the wolfspeed manufacturer. The nonlinear capacitor parameters and the original parameters of the analog current source before simplification are obtained from this model.

[0206] A double pulse test platform was built to test the switching transient process, and the results were compared under different input voltages and input currents. One set of simulation parameter settings is shown in the following table.

[0207] Table 2 Double pulse test simulation parameters

[0208] Input voltage 600V Gate resistance 10Ω Input Current 40A Stray inductance 40nH Shutoff voltage -5V On-state voltage 16V

[0209] Figure 4 For this parameter setting, the drain current waveform comparison is as follows: Figure 7 is the gate voltage contrast, Fig.10 The figure shows the comparison of drain-source voltage waveforms. It can be seen that the multiple piecewise linear model has a good simulation effect on a long time scale. Figure 5 , Figure 8 , Fig.11 For the current and voltage response of the turn-on transient, it can be found that during the turn-on process, the multiple piecewise linear model can accurately reflect transient characteristics such as the Miller platform and the drain current overshoot caused by the reverse recovery of the diode. Figure 6 , Fig. 9 , Fig.12 For the current and voltage response of the shutdown transient, it can be found that the multiple piecewise linear model of the shutdown process can accurately reflect the drain-source voltage overshoot oscillation phenomenon. At the nanosecond time scale, the model established by the present invention can still be almost consistent with the simulation results of commercial-grade simulation software.

[0210] The double pulse test was carried out under the above conditions, and the parameters characterizing the transient characteristics of the switch are shown in Table 3 and Table 4. The rise time is from 10% of the final gate voltage value to the drain-source voltage falling to 90% of its initial value. The turn-off delay time is the time it takes for the drain-source voltage to drop from 90% of the initial value to 10%. The fall time is from the time when the gate voltage drops to 90% of its initial value to the time when the drain-source voltage rises to 10% of its final value. It is the time for the drain-source voltage to rise from 10% to 90%. It is the energy loss generated during the transient process of the switch tube turning on, that is, the integral of the product of the drain-source voltage and the drain current over time from 10% of the final gate voltage value to the drain-source voltage dropping to 2% of its initial value. It is the energy loss generated in the transient process of the switch tube turning off, that is, the integration of the product of the drain-source voltage and the drain current over time from 90% of the initial gate voltage to 2% of its initial value. It can be found from the table that the relative errors of all relevant dynamic parameters are less than 10%.

[0211] Table 3 Statistics of dynamic parameter errors

[0212] Dynamic parameter names LTspice Simulation Results Piecewise linear method simulation results Absolute error Relative error Turn-on delay time 2.966ns 3.250ns 0.284ns 9.58% Rise time 1.257ns 1.175ns -0.082ns 6.55% Turn-off delay time 9.551ns 9.750ns 0.199ns 2.08% Fall time 4.501ns 4.500ns -0.001ns 0.02%

[0213] Table 4 Statistics of switching energy loss errors

[0214] Dynamic parameter names LTspice Simulation Results Piecewise linear method simulation results Absolute error Relative error Turn-on loss 0.8524mJ 0.8363mJ -0.0161mJ 1.88% Turn-off loss 0.1715mJ 0.1719mJ 0.0004mJ 0.19%

[0215] The multiple piecewise linear switching transient model constructed by the method of the present invention is applied to Fig.13 In the buck converter shown in FIG. 1 , the simulation parameter settings are shown in the following table.

[0216] Table 5 Buck converter test simulation parameters

[0217] enter 15V inductance 5mH load 2.5Ω capacitance 100μF Switching frequency 200kHz Duty Cycle 0.5

[0218] Fig.14 This is a comparison chart of the output voltage waveform of the buck converter. Fig.15 The figure is a comparison of the converter inductor current waveforms. It can be seen from the figure that the multiple piecewise linear model can be used for converter-level calculations.

[0219] In summary, the method of the present invention can accurately simulate the switching transient of SiC MOSFET and reduce the computational cost of solving the power semiconductor model due to its non-iterative nature; the method of the present invention is used in buck converter simulation to well simulate the output characteristics of the buck converter and the transient characteristics of the switch.

[0220] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any way. Any simple modification, change and equivalent structural change made to the above embodiment based on the technical essence of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A method for solving transient multi-piecewise linear modeling of power electronic converter switches, characterized in that: The following steps are involved: A1. Determine the calculation unit and use the half-bridge converter as the calculation unit; A2. Build a behavioral model based on the transient process of the switch module; A3. According to the state of the switch tube and the diode, the switching transient is divided into three different stages and equivalent circuits are established for each of them; in each stage of the switching transient, the diode reverse process is selected to simulate the inductor current i Lrr , drain current i ds , gate voltage v gs , drain-source voltage v ds Four variables are used as state variables, and the three stages specifically include: Phase 1: The diode is forward conducting. The ideal diode in the diode model is equivalent to a voltage source and a resistor in series, where the voltage value of the voltage source is the diode forward voltage drop V d_on The resistance value is the forward resistance of the diode R d_on , write the current equation for the diode model, there is a relationship: The input voltage source V cc The Kirchhoff voltage equation can be written as follows for the main circuit consisting of a diode and a switch: Among them, V ds_off is the drain-source voltage in the steady state when the switch is turned off. Under this equivalent condition, its expression is V ds_off =V d_on +I L R d_on +V cc ; The gate voltage of the switch tube is lower than the threshold voltage V th , in the cut-off region, the analog current source I used to simulate the dynamic characteristics of the switch tube mos =0, at this stage, writing the current equation for the gate can be obtained: Among them, V g is the gate control voltage, R g is the gate resistance; Writing the current equation for the source gives: Phase 2: The diode is forward-conducting, the gate voltage of the switch tube is higher than the threshold voltage, and the analog current source I mos With gate voltage v gs and drain-source voltage ν ds Related, in accordance with the functional relationship: I mos =f(v ds , v gs ), write the current equation for the diode model at this stage, and there is a relationship: Writing Kirchhoff's voltage equation for the main circuit of this stage gives: Writing the current equation for the gate gives: Writing the current equation for the source gives: Stage 3: The diode is reversely turned off, and the ideal diode in the diode model is equivalent to an open circuit; the gate voltage of the switch tube is higher than the threshold voltage, and the simulated current source is related to the gate voltage and the drain-source voltage, which conforms to the functional relationship: I mos =f(v ds ,v gs ); At this stage, since the ideal diode is equivalent to an open circuit, the reverse recovery simulation resistance and simulation inductance in the diode model conform to the relationship: Diode module current i d Conforms to the relationship: in d =Kv Lrr =-KR rr in Lrr , The relationship between the diode current and the drain current is as follows: i ds =I L -i d , Writing the current equation for the gate in the equivalent circuit of the switch tube gives: Writing the current equation for the source gives: A4. Use multiple piecewise linear methods to solve the equivalent circuits of each stage non-iteratively. A5. Apply multiple piecewise linear models to solve the circuit of buck converter; Among them, I L Indicates input current, L s The equivalent inductance of stray inductance, C ds The equivalent variable capacitance of the parasitic capacitance between drain and source, C dg The equivalent variable capacitance of the parasitic capacitance between drain and gate, C gs Represents the equivalent capacitance of the parasitic capacitance between the gate and the source, L rr Represents the inductor simulating the reverse recovery characteristics, R rr Represents the resistance simulating the reverse recovery characteristic, Kv Lrr represents a controlled current source simulating reverse recovery characteristics, K represents a controlled current source coefficient; The behavioral model built in step A2 can accurately represent the output characteristics and transfer characteristics of the cut-off region and the non-cut-off region during the transient process of the switch tube with low complexity, and can accurately represent the reverse recovery characteristics of the diode. The model of the switch tube SiC MOSFET consists of the equivalent inductance L of the stray inductance s , the equivalent variable capacitance C of the parasitic capacitance between drain and source ds , the equivalent variable capacitance C of the parasitic capacitance between the drain and the gate dg , the equivalent capacitance C of the parasitic capacitance between the gate and the source gs And a controlled current source I used to simulate the dynamic characteristics of the switch tube mos The diode model consists of an ideal diode and an inductor L that simulates the reverse recovery characteristics. rr , resistor R rr , controlled current source Kv Lrr composition; The specific steps of performing non-iterative solution of the equivalent circuit in step A4 include: Step B1, loading the calculation result of the previous step; Step B2: Determine which stage is in the process of opening, the gate control voltage V g >0, the switch tube switches from the cut-off state to the on state, and the diode switches from the on state to the cut-off state. First, determine whether the switch tube is in the cut-off region. The judgment basis is whether the gate voltage reaches the threshold voltage: If v gs ≤V th , then it is in stage 1, the analog current source I mos =0, then judge the diode state, the opening process adopts the current judgment method: if the current I flowing through the diode model d Greater than the controlled current source current, that is, I d ≥Kv Lrr , the ideal diode in the diode model remains on and is in stage two; otherwise, the ideal diode is disconnected and is in stage three. During the shutdown process, the gate control voltage V g ≤0, the switch tube switches from the on state to the off state, and the diode switches from the off state to the on state. First, the diode state is judged, and the shutdown process adopts the voltage judgment method: if v ds ≤V cc +V d_on , the ideal diode in the diode model withstands the reverse voltage and remains disconnected, which is in stage three; otherwise, the ideal diode is turned on, and then the switch state is judged. If v gs >V th , the switch tube has not yet reached the cut-off zone and is in stage 2; otherwise, it switches to the cut-off state and is in stage 1; Step B3: Calculate and update the value of equivalent parasitic capacitance. C gs is a constant value, C dg and C ds They are all variable capacitors, and their values ​​change with the voltage across the capacitor, which conforms to the functional relationship C ds =g(v ds ), C g =h(v dg ), where v dg =v ds -v gs ; Step B4: Use a lookup table to determine the value of the analog current source. The specific implementation method is to first linearize the analog current source and convert the original complex nonlinear function I mos =f(v ds , v gs ) is represented by I mos =k0+k1v gs +k2v ds In the form of, k0, k1 and k2 are linearization coefficients. According to the principle of multivariate function linearization, v in this form gs =a,v ds = b, the linearization coefficient k1 is the original complex nonlinear function at v gs =a,v ds = b This point is opposite to v gs The partial derivative of v ds The partial derivative of k0 is a constant term, and its expression is k0=f(a,b)-k1·a-k2·b, then choose the appropriate v ds and v gs Segment distance, the function I mos =f(v ds , v gs ) is divided into multiple planes, and the linearization results are stored in the form of a two-dimensional table. ds , v gs ) is used as an index to determine the values ​​of k0, k1, and k2; Step B5: Write the correct state equation according to different stages to solve the circuit and get the matrix form of the state equation. Among them, X is the state variable vector, u is the input vector, coefficient matrix A is the state matrix, and coefficient matrix B is the control matrix. First, define the coefficient For stage 1, the matrix form of the state equation can be expressed as: Among them, the diode reverse process simulates the inductor current i Lrr , drain current i ds , gate voltage v gs , drain-source voltage v ds Four variables are state variables; The expression of the coefficient matrix is: State Matrix Control Matrix For stage 2, the matrix form of the state equation can be expressed as: The expression of the coefficient matrix is: State Matrix Control Matrix Where K is the controlled current source coefficient, k3 is the linearization coefficient of the current source in the switch tube equivalent circuit; for stage three, the diode state equation is: The calculation method of the switch tube drain current is: and ds =I L +KR rr and Lrr , The matrix form of the switch state equation can be expressed as: in, The state equations obtained for different stages are Perform numerical integration and use the backward Euler method. The time step used in the calculation process is h. The state variable value of the current step is expressed as n+1 =(I-hA) -1 X n +h(I-hA) - 1 Bu updates the calculation results, where I is the unit matrix, X n is the state variable vector of the current step, X n+1 is the state variable vector of the next step length; Step B6: Update the network with the solution result, end the current step length, and wait for the next step length to start.

2. The method for solving the transient multi-piecewise linear modeling of a power electronic converter switch according to claim 1, characterized in that: In step A1, the half-bridge converter includes two upper and lower SiC MOSFET / diode pairs. In the switching event, the upper and lower SiC MOSFETs are not turned on at the same time. It can be simplified as the upper switch tube and the lower diode are always in the off state, which is equivalent to an open circuit. Only the upper diode and the lower switch tube are affected during the switching process, and the input voltage V of the calculation unit is calculated during the transient calculation process. cc and input current I L Considered as a constant value.

3. The method for solving the transient multi-piecewise linear modeling of a power electronic converter switch according to claim 1, characterized in that: In step A5, the multiple piecewise linear model is applied to the circuit solution of the buck converter. The device-level simulation of the power electronic converter in the buck converter means that the switching devices in the buck converter are solved by the multiple piecewise linear method, and the other linear parts of the circuit are solved by the backward Euler method.

4. A power electronic converter switch transient multiple piecewise linear modeling system based on the power electronic converter switch transient multiple piecewise linear modeling solution method according to any one of claims 1 to 3, characterized in that: The system includes: a calculation unit module, used to determine to use a half-bridge converter as a calculation unit; Behavior model building module, used to build a behavior model according to the transient process of the switch tube module; An equivalent circuit establishment module is used to divide the switching transient state into three different stages according to the states of the switch tube and the diode and establish equivalent circuits for each of them; The non-iterative solution module is used to perform non-iterative solutions to the equivalent circuits of each stage respectively by using a multiple piecewise linear method; and the application module is used to apply the multiple piecewise linear model to the circuit solution of the buck converter.

5. The power electronic converter switch transient multiple piecewise linear modeling system according to claim 4, characterized in that: The computing unit module comprises: The half-bridge converter includes two upper and lower SiC MOSFET / diode pairs. The upper switch tube and the lower diode are simplified to be in the off state, which is equivalent to an open circuit. Only the upper diode and the lower switch tube are affected during the switching process, and the input voltage and input current of the calculation unit are regarded as constants during the transient calculation process.

6. The power electronic converter switch transient multiple piecewise linear modeling system according to claim 4, characterized in that: The behavioral model building module can accurately represent the output characteristics and transfer characteristics of the cut-off region and the non-cut-off region in the transient process of the switching tube, as well as the reverse recovery characteristics of the diode with low complexity. The model of the switching tube SiC MOSFET consists of an equivalent inductance of stray inductance, an equivalent variable capacitance of the parasitic capacitance between the drain and the source, an equivalent variable capacitance of the parasitic capacitance between the drain and the gate, an equivalent capacitance of the parasitic capacitance between the gate and the source, and a controlled current source for simulating the dynamic characteristics of the switching tube. The diode model consists of an ideal diode and an inductor, a resistor, and a controlled current source for simulating the reverse recovery characteristics.

7. The power electronic converter switch transient multiple piecewise linear modeling system according to claim 4, characterized in that: The equivalent circuit establishment module divides the switching transient into three stages. In each stage, four variables, namely, the inductor current, drain current, gate voltage, and drain-source voltage, are selected to simulate the diode reverse process as state variables. The equivalent circuits of stage one, stage two, and stage three are established respectively. The correct state equations are written for different stages to solve the circuits and obtain the matrix form of the state equations. Among them, X is the state variable vector, u is the input vector, coefficient matrix A is the state matrix, coefficient matrix B is the control matrix, and the numerical integration is solved by the backward Euler method. The time step used in the calculation process is h, and the state variable value of the current step is expressed as n+1 =(I-hA) -1 X n +h(I-hA) -1 Bu updates the calculation results and uses the solution results to update the network, ending the current step length and waiting for the next step length to start; where I is the unit matrix, X n is the state variable vector of the current step, X n+1 is the state variable vector of the next step length.