A method and system for testing the response rate of a power MOS tube

Abstract

The invention discloses a response rate test method and system for a power MOS tube, and relates to the technical field of power semiconductor device testing, comprising: setting working parameters of the MOS tube to simulate its working environment in actual application; capturing charge distribution and thermal effect generated by the MOS tube in a switching process by an optical microinterferometer to obtain initial time domain response data; converting the time domain response data into frequency domain data to obtain frequency response characteristics; combining the time domain response data and the frequency domain data to comprehensively evaluate the response rate of the MOS tube under different frequencies and different working conditions, and generating a dynamic response rate report; constructing a response rate prediction model according to the dynamic response rate report, training and optimizing the response rate prediction model through historical data, inputting the working parameters of the MOS tube into the response rate prediction model, and predicting the response rate of the MOS tube in a future period of time; comparing the actual test results with the predicted results, and dynamically adjusting the working parameters of the MOS tube according to the error.

Description

Technical Field

[0001] The present invention relates to the technical field of power semiconductor device testing, and in particular to a method and system for testing the response rate of a power MOS tube. Background Art

[0002] As an important power semiconductor device, power MOS tube is widely used in the switching circuit of power electronic system. Because of its advantages of high switching speed, low on-resistance and high power efficiency, since the 1980s, when power MOS tube gradually replaced bipolar transistor, its technology has developed rapidly and has become one of the indispensable components in modern power electronic equipment. With the advancement of science and technology, the application scenarios of power MOS tube are constantly expanding, from household appliances to industrial control, to new energy vehicles and high-performance computing. The response rate of power MOS tube directly affects the switching efficiency and overall working performance of the circuit. Therefore, how to accurately evaluate and optimize the response rate of power MOS tube in practical applications has become a key issue in the design, testing and optimization of power devices. The traditional power MOS tube test method mainly relies on electrical characteristics test. Although it can obtain certain response data, it cannot fully reflect the dynamic response rate of the device under complex working conditions such as high frequency and high voltage, especially the impact of nonlinear charge distribution and thermal effect on the response rate generated during the switching process.

[0003] The existing technology often has the following deficiencies when testing the response rate of power MOS tubes. First, most traditional electrical testing methods can only obtain current and voltage change data in the time domain, and it is difficult to capture the dynamic changes in charge distribution and its thermal effects during the high-speed switching process of the MOS tube. Due to the nonlinear changes in the gate voltage, drain current and source voltage of the power MOS tube during the high-speed switching process, these physical phenomena have a significant impact on the response rate of the MOS tube. Simply relying on electrical testing cannot fully evaluate these factors. Secondly, the existing testing methods lack analysis of frequency domain response and cannot fully reveal the dynamic performance of the power MOS tube at different frequencies and its response rate under high-frequency switching conditions. In addition, traditional testing methods usually only provide test results under static or limited conditions, and it is difficult to train and predict models through historical data. There is a lack of comprehensive prediction capabilities for the response rate of the MOS tube under future working conditions. These deficiencies make it difficult for engineers to obtain comprehensive dynamic performance data when designing and optimizing power MOS tubes, which in turn limits the optimization and improvement of the response rate. Summary of the invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides a response rate testing method for a power MOS tube to solve the problem that the existing testing methods lack analysis of frequency domain response and cannot fully reveal the dynamic performance of the power MOS tube at different frequencies and its response rate under high-frequency switching conditions.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, the present invention provides a method for testing the response rate of a power MOS tube, which comprises:

[0008] Set the working parameters of the MOS tube to simulate its working environment in actual applications;

[0009] The charge distribution and thermal effects generated by the MOS tube during the switching process are captured by an optical micro-interferometer to obtain the initial time domain response data;

[0010] Convert the time domain response data into frequency domain data to obtain the frequency response characteristics;

[0011] Combine time domain response data and frequency domain data to comprehensively evaluate the response rate of MOS tubes at different frequencies and working conditions, and generate a dynamic response rate report;

[0012] Build a response rate prediction model based on the dynamic response rate report, train and optimize the response rate prediction model through historical data, input the working parameters of the MOS tube into the response rate prediction model, and predict the response rate of the MOS tube in the future;

[0013] Compare the actual test results with the predicted results, and dynamically adjust the operating parameters of the MOS tube according to the error.

[0014] As a preferred solution of the response rate test method of the power MOS tube of the present invention, the setting of the working parameters of the MOS tube to simulate its working environment in actual application includes:

[0015] Set the voltage, current and frequency of the MOS tube;

[0016] Collect the gate-source voltage and drain-source voltage of the MOS tube and calculate the working current of the MOS tube. The expression is:

[0017]

[0018] in, is the drain current, is the drain-source voltage, is the on-resistance of the MOS tube, represents the base of natural logarithms, is the gate-source voltage, Indicates the threshold voltage of the MOS tube;

[0019] The switching frequency of the MOS tube is calculated by the time between the off state and the on state of the MOS tube. The expression is:

[0020]

[0021] in, represents the switching frequency, Indicates the time it takes for the MOS tube to switch from the off state to the on state. Indicates the time it takes for the MOS tube to switch from the on state to the off state.

[0022] As a preferred solution of the response rate test method of the power MOS tube of the present invention, wherein: the charge distribution and thermal effect generated by the MOS tube during the switching process are captured by an optical micro-interferometer to obtain the initial time domain response data, the specific steps are:

[0023] An optical micro-interferometer is used to illuminate the area from the gate to the source of the MOS tube to measure the deformation and thermal effects of the area;

[0024] After the working conditions of the MOS tube are set, the optical micro-interferometer is started to capture the changes in charge distribution and thermal effects generated by the MOS tube during the switching process;

[0025] Calculate the charge distribution change of the MOS tube during the switching process, the expression is:

[0026]

[0027] in, Indicates at time The charge distribution of represents the maximum accumulation of charge, represents the time constant of the change of charge distribution;

[0028] Calculate the change of thermal effect of MOS tube during switching process, the expression is:

[0029]

[0030] in, Indicates at time The physical length change, that is, the expansion or contraction of the surface of the MOS tube due to thermal effects, Indicates the initial length of the MOS tube surface before work. It indicates the rate of change of the linear dimension of a material under unit temperature change. Indicates time The temperature change;

[0031] The response rate of the MOS tube is calculated based on the change in charge distribution and thermal effect. The expression is:

[0032]

[0033] in, Indicates that the MOS tube is in time The response rate, represents the weight of the charge distribution change, The weight representing the change in thermal effect.

[0034] As a preferred solution of the response rate test method of the power MOS tube of the present invention, wherein: the time domain response data is converted into frequency domain data to obtain the frequency response characteristics, specifically,

[0035] Using Fast Fourier Transform FFT Convert to frequency domain signal ;

[0036] By analyzing the frequency domain signal The frequency response characteristics are obtained, and the amplitude and phase calculation formulas of the frequency domain signal are:

[0037]

[0038]

[0039] in, Indicates the switching frequency The signal amplitude at Indicates the switching frequency The signal phase at Represents the argument function of a complex number.

[0040] As a preferred solution of the response rate test method of the power MOS tube of the present invention, wherein: the time domain response data and the frequency domain data are combined to comprehensively evaluate the response rate of the MOS tube under different frequencies and different working conditions, and generate a dynamic response rate report, specifically,

[0041] Combine the response rate of the MOS tube, the amplitude and phase of the frequency domain signal;

[0042] The response rate of the MOS tube at different frequencies and working conditions is comprehensively evaluated to obtain the comprehensive response rate, which is expressed as:

[0043]

[0044] in, Indicates time, is the comprehensive response rate, represents the weight of the time domain response data, Represents the weight of frequency domain response data;

[0045] According to the overall response rate Generates a dynamic response rate report that includes a comprehensive response rate curve, frequency response characteristic analysis, and time domain response characteristic analysis.

[0046] As a preferred solution of the response rate test method of the power MOS tube of the present invention, wherein: the response rate prediction model is constructed according to the dynamic response rate report, and the response rate prediction model is trained and optimized through historical data, and the specific steps are:

[0047] Build a response rate prediction model based on the generated dynamic response rate report, select the neural network model as the response rate prediction model, and select the multi-layer perceptron MLP as the basic structure of the neural network;

[0048] The gate-source voltage, drain-source voltage, switching frequency, on-resistance, time domain response rate, frequency domain amplitude and frequency domain phase of the MOS tube are used as inputs of the response rate prediction model;

[0049] The multi-layer formula for predicting response rate is:

[0050]

[0051]

[0052]

[0053] in, and They correspond to the weights of the gate-source voltage, drain-source voltage, switching frequency, on-resistance, time domain response rate, frequency domain amplitude and frequency domain phase of the MOS tube, respectively. represents the hyperbolic tangent function, represents the intermediate features passed from the first layer, , and represents the parameters with training, and They represent the first and second nonlinear mapping functions of the response rate prediction model, represents the predicted response rate;

[0054] The inputting of the working parameters of the MOS tube into the response rate prediction model to predict the response rate of the MOS tube in the future period of time refers to combining the nonlinear mapping functions of the first layer and the second layer to obtain ;

[0055] Select mean square error MSE as the loss function to measure the gap between the predicted value and the true value;

[0056] A large amount of historical data is extracted from the dynamic response rate report, and the data set is divided into a training set and a test set. The training set data is input into the response rate prediction model as training, and the network weights are continuously adjusted through the back propagation algorithm. In each iteration, the error between the predicted response rate and the actual response rate is calculated, and the loss function is minimized through the gradient descent method;

[0057] The error between the predicted results and the actual test results is calculated using the test set.

[0058] As a preferred solution of the response rate test method of the power MOS tube of the present invention, the actual test result is compared with the predicted result, and the working parameters of the MOS tube are dynamically adjusted according to the error, specifically,

[0059] Use optical micro-interferometer and related sensors to obtain the response rate data of MOS tube under the current working parameter conditions, and set the actual response rate collected to ;

[0060] Use mean square error (MSE) as the loss function, the error between the predicted result and the actual test result (MSE);

[0061] Dynamically update each parameter value according to the error value MSE, the expression is:

[0062]

[0063] in, Indicates the working parameters to be adjusted, represents the number of iterations, represents the learning rate, represents partial derivative.

[0064] In a second aspect, the present invention provides a response rate test system for a power MOS tube, comprising: a setting module, a collection module, an analysis module, a prediction model module and an adjustment module.

[0065] The setting module is used to set and control the working parameters of the MOS tube to simulate its working environment in actual applications;

[0066] The acquisition module captures the charge distribution changes and thermal effects generated by the MOS tube during the switching process through an optical micro-interferometer, obtains initial time domain response data, and converts it into frequency domain data;

[0067] The analysis module is used to comprehensively process the time domain and frequency domain response data, evaluate the response rate of the MOS tube under different frequencies and working conditions, and generate a dynamic response rate report;

[0068] The prediction model module trains and optimizes the response rate prediction model based on historical data and the generated dynamic response rate report to predict the response rate of the future MOS tube;

[0069] The adjustment module is used to compare the actual test results with the predicted results, and dynamically adjust the working parameters of the MOS tube through the error feedback mechanism mean square error MSE to optimize its response rate.

[0070] In a third aspect, the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, any step of the response rate testing method for a power MOS tube as described in the first aspect of the present invention is implemented.

[0071] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, any step of the method for testing the response rate of a power MOS tube as described in the first aspect of the present invention is implemented.

[0072] The beneficial effects of the present invention are as follows: the present invention provides a power MOS tube response rate testing method based on an optical microinterferometer, which can effectively capture the dynamic changes of charge distribution and thermal effects of the MOS tube during the switching process, and provide more accurate time domain response data. At the same time, the time domain data is converted into frequency domain data through fast Fourier transform FFT, and the response rate of the MOS tube at different frequencies is comprehensively evaluated. The method can not only generate a dynamic response rate report, but also dynamically adjust the working parameters of the MOS tube by training and optimizing the response rate prediction model through historical data, combining the comparison between the actual test results and the prediction results, thereby realizing the optimization and improvement of the response rate of the MOS tube. The method invented by our party has high accuracy and dynamic adaptability, especially under complex working conditions of high frequency and high voltage, and can significantly improve the response rate evaluation and optimization effect of the power MOS tube. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0074] Figure 1 This is a flow chart of the response rate testing method of the power MOS tube in Example 1.

[0075] Figure 2 Schematic diagram of the response rate test system of the power MOS tube in Example 1. DETAILED DESCRIPTION

[0076] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0077] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0078] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0079] Example 1, reference Figure 1 and Figure 2 , which is the first embodiment of the present invention, provides a method for testing the response rate of a power MOS tube, comprising the following steps:

[0080] S1. Set the working parameters of the MOS tube to simulate its working environment in actual application;

[0081] Setting the voltage, current and frequency of the MOS tube. This step accurately simulates the working conditions of the MOS tube in actual applications by setting the voltage, current and frequency. Compared with the single electrical parameter test method in the prior art, the present invention controls multiple working parameters to make the test conditions closer to the actual application environment. This setting can more comprehensively reflect the performance of the MOS tube under complex working conditions, which helps to improve the accuracy and repeatability of the test results.

[0082] The gate-source voltage and drain-source voltage of the MOS tube are collected, and the gate voltage is adjusted to control the drain current, thereby adjusting the working state of the MOS tube and calculating the working current of the MOS tube. The expression is:

[0083]

[0084] in, is the drain current, indicating the working current of the MOS tube. is the drain-source voltage, which indicates the voltage of the MOS tube when it is working. is the on-resistance of the MOS tube, represents the base of natural logarithms, is the gate-source voltage, Indicates the threshold voltage of the MOS tube, which indicates the minimum gate voltage of the MOS tube from off to on;

[0085] This formula not only takes into account the linear relationship between voltage and resistance, but also introduces an exponential decay term , which reflects the nonlinear control of the gate voltage on the drain current during the switching process of the MOS tube. This calculation method can more accurately capture the dynamic working state of the MOS tube, especially the transient changes of the current during the switching process;

[0086] The switching frequency of the MOS tube is calculated by the time between the off state and the on state of the MOS tube. The expression is:

[0087]

[0088] in, represents the switching frequency, Indicates the time it takes for the MOS tube to switch from the off state to the on state. Indicates the time it takes for the MOS tube to switch from the on state to the off state;

[0089] The present invention can accurately calculate the switching frequency of the MOS tube. Compared with the method of measuring frequency by a timer or an oscilloscope in the prior art, the present invention directly calculates the frequency based on the accurate measurement of the switching time, which can better reflect the actual operating frequency of the MOS tube. The advantage of this method is that it not only improves the accuracy of frequency measurement, but also can capture the dynamic performance of the MOS tube under high-frequency switching, and is particularly suitable for high-speed switching application scenarios, such as power converters in electric vehicles or high-frequency power supplies.

[0090] S2. Capture the charge distribution and thermal effect generated by the MOS tube during the switching process through an optical micro-interferometer to obtain the initial time domain response data;

[0091] An optical micro-interferometer is used to illuminate the area from the gate to the source of the MOS tube, and the deformation and thermal effect of the area are measured, thereby providing a high-precision physical change signal for the acquisition of time domain response data;

[0092] After the working conditions of the MOS tube are set, the optical micro-interferometer is started to capture the changes in charge distribution and thermal effects generated by the MOS tube during the switching process. These changes will cause the surface deformation of the MOS tube, which is recorded through the changes in interference fringes of the optical micro-interferometer.

[0093] During the switching process of the MOS tube, the change of the gate voltage will cause the rearrangement of the charge distribution between the source and the drain. The power MOS tube will generate heat during the high-frequency switching process. This thermal effect will cause a slight expansion or contraction of the material, which in turn affects the change of the interference fringes. The optical micro-interferometer detects these thermal effects and records them as time domain response data.

[0094] The innovation of this step is to use the high-precision tool of optical micro-interferometer to capture the charge distribution and thermal effect generated by the MOS tube during the switching process. Traditional electrical testing methods usually cannot directly observe the charge dynamics and the accompanying thermal effects inside the MOS tube, while the optical micro-interferometer can accurately detect these tiny physical changes through the changes in interference fringes. This detection method not only provides extremely high spatial resolution, but also can perform real-time monitoring without interfering with the normal operation of the MOS tube, greatly improving the accuracy of the test;

[0095] Calculate the charge distribution change of the MOS tube during the switching process, the expression is:

[0096]

[0097] in, Indicates at time The charge distribution of the MOS tube is the accumulated charge during the switching process. represents the maximum accumulation of charge, represents the time constant of the change of charge distribution;

[0098] Calculate the change of thermal effect of MOS tube during switching process, the expression is:

[0099]

[0100] in, Indicates at time The physical length change, that is, the expansion or contraction of the surface of the MOS tube due to thermal effects, Indicates the initial length of the MOS tube surface before work. It indicates the rate of change of the linear dimension of a material under unit temperature change. Indicates time The temperature change;

[0101] The response rate of the MOS tube is calculated based on the change in charge distribution and thermal effect. The expression is:

[0102]

[0103] in, Indicates that the MOS tube is in time The response rate, i.e. the time domain response data, represents the weight of the charge distribution change, The weight representing the change of thermal effect;

[0104] This comprehensive calculation method is innovative because it takes two different physical phenomena (charge distribution and thermal effect) into account and can fully reflect the dynamic response characteristics of the MOS tube. Compared with the existing technology that analyzes electrical or thermal effects separately, the present invention introduces weight coefficients to and , it can flexibly adjust the impact of charge distribution and thermal effects on response rate according to actual application scenarios, so as to better adapt to testing requirements under different frequencies and powers.

[0105] S3, converting the time domain response data into frequency domain data to obtain frequency response characteristics;

[0106] Using Fast Fourier Transform FFT Convert to frequency domain signal ;

[0107] By analyzing the frequency domain signal The frequency response characteristics, namely the amplitude and phase, are obtained. The amplitude and phase calculation formulas of the frequency domain signal are as follows:

[0108]

[0109]

[0110] in, Indicates the switching frequency The signal amplitude at Indicates the switching frequency The signal phase at Represents the argument function of a complex number;

[0111] By converting time domain data into frequency domain data and analyzing the amplitude and phase of the signal in combination with the FFT algorithm, the present invention can comprehensively and quickly evaluate the frequency response characteristics of the MOS tube. The comprehensive analysis of the amplitude and phase provides more abundant and accurate information for the performance evaluation of the MOS tube. Especially in high-frequency applications, these key steps of the present invention can significantly improve the insight into the dynamic behavior of the MOS tube and help optimize its design and application.

[0112] The present invention can analyze the dynamic performance of the MOS tube from a frequency perspective. Compared with the limitation of the prior art that only analyzes through time domain data, frequency domain analysis can reveal the response characteristics at different frequencies. In particular, under high-frequency working conditions, frequency domain analysis can more comprehensively reflect the dynamic response behavior of the MOS tube. For example, in high-frequency application scenarios, the frequency response characteristics are particularly critical for evaluating the reliability and performance of the MOS tube. Through frequency domain analysis, the loss, stability, resonance and other characteristics of the MOS tube at different frequencies can be accurately evaluated, thereby providing an important basis for optimizing the design and application of the MOS tube.

[0113] S4. Combine time domain response data and frequency domain data to comprehensively evaluate the response rate of the MOS tube under different frequencies and working conditions, and generate a dynamic response rate report;

[0114] Combine the response rate of the MOS tube, the amplitude and phase of the frequency domain signal;

[0115] The response rate of the MOS tube at different frequencies and working conditions is comprehensively evaluated to obtain the comprehensive response rate, which is expressed as:

[0116]

[0117] in, Indicates time, It is the comprehensive response rate, which indicates the response rate of the power MOS tube at a specific time and frequency. represents the weight of the time domain response data, Represents the weight of frequency domain response data;

[0118] According to the overall response rate Generate dynamic response rate report including comprehensive response rate curve, frequency response characteristic analysis and time domain response characteristic analysis;

[0119] Comprehensive response rate curve: displays the response rate curve of MOS tube under different frequencies, different gate voltages, and different drain currents. The horizontal axis of the curve is frequency, and the vertical axis is response rate.

[0120] Frequency response characteristic analysis: Displays the amplitude and phase information in the frequency domain, and can identify the important frequency range that affects the dynamic performance of the MOS tube;

[0121] Time domain response characteristics analysis: The changes in charge distribution and thermal effects of the MOS tube during the switching process are analyzed through time domain data. This part of the data can reflect the charge accumulation and thermal effects of the MOS tube at different time points, and further reveal the dynamic performance;

[0122] The generated dynamic response rate report provides engineers with a comprehensive tool to intuitively understand the response performance of the device during the MOS tube design, optimization and testing process. By analyzing the report, engineers can identify under what operating conditions the MOS tube responds fastest or slowest and optimize its operating parameters.

[0123] S5. Build a response rate prediction model based on the dynamic response rate report, train and optimize the response rate prediction model through historical data, input the working parameters of the MOS tube into the response rate prediction model, and predict the response rate of the MOS tube in the future;

[0124] Build a response rate prediction model based on the generated dynamic response rate report, select the neural network model as the response rate prediction model, and select the multi-layer perceptron MLP as the basic structure of the neural network. MLP consists of multiple layers of neurons and can capture the nonlinear relationship between input features.

[0125] The gate-source voltage, drain-source voltage, switching frequency, on-resistance, time domain response rate, frequency domain amplitude and frequency domain phase of the MOS tube are used as inputs of the response rate prediction model;

[0126] The multi-layer formula for predicting response rate is:

[0127]

[0128]

[0129]

[0130] in, , , , , , and They correspond to the weights of the gate-source voltage, drain-source voltage, switching frequency, on-resistance, time domain response rate, frequency domain amplitude and frequency domain phase of the MOS tube, respectively. represents the hyperbolic tangent function, represents the intermediate features passed from the first layer, , and represents the parameters with training, and They represent the first and second nonlinear mapping functions of the response rate prediction model, represents the predicted response rate;

[0131] Inputting the working parameters of the MOS tube into the response rate prediction model, predicting the response rate of the MOS tube in the future period of time means combining the nonlinear mapping functions of the first layer and the second layer to obtain ;

[0132] Select mean square error MSE as the loss function to measure the gap between the predicted value and the true value;

[0133] A large amount of historical data is extracted from the dynamic response rate report, and the data set is divided into a training set and a test set. The training set data is input into the response rate prediction model as training, and the network weights are continuously adjusted through the back propagation algorithm. In each iteration, the error between the predicted response rate and the actual response rate is calculated, and the loss function is minimized through the gradient descent method;

[0134] Use the test set to calculate the error between the predicted results and the actual test results;

[0135] The first layer of nonlinear mapping function The role of is to map the input features to an intermediate feature space through nonlinear combination;

[0136] The second layer nonlinear mapping function The function of is to further process the intermediate features and output the final response rate prediction result;

[0137] This formula uses multi-layer nonlinear mapping, combining hyperbolic tangent function, sigmoid function and exponential function, which can accurately capture the complex changes of MOS tube response rate under different working conditions. It combines the nonlinear characteristics of neural network, can be trained through historical data, and gradually optimize weight parameters to ensure accurate prediction results under different working conditions. This prediction model can be applied to the dynamic adjustment and optimization of MOS tubes. By predicting the response rate in real time, it helps engineers adjust working conditions and improve the performance and stability of MOS tubes.

[0138] S6. Compare the actual test results with the predicted results, and dynamically adjust the working parameters of the MOS tube according to the error;

[0139] Use optical micro-interferometer and related sensors to obtain the response rate data of MOS tube under the current working parameter conditions, and set the actual response rate collected to ;

[0140] Using mean square error MSE as the loss function, the error MSE between the predicted result and the actual test result is expressed as:

[0141]

[0142] in, represents the error value, represents the number of data points, Indicates the actual response rate, represents the predicted response rate, is an index variable with a value range of 1 to 7, corresponding to the gate-source voltage, drain-source voltage, switching frequency, on-resistance, time domain response rate, frequency domain amplitude and frequency domain phase of the MOS tube respectively;

[0143] According to the error value Dynamically update each parameter value, the expression is:

[0144]

[0145] in, Indicates the operating parameters to be adjusted, including the gate-source voltage, drain-source voltage, switching frequency, on-resistance, time domain response rate, frequency domain amplitude and frequency domain phase of the MOS tube. represents the number of iterations, represents the learning rate, represents partial derivative.

[0146] This embodiment also provides a power MOS tube response rate test system, including: a setting module, a collection module, an analysis module, a prediction model module and an adjustment module.

[0147] The setting module is used to set and control the working parameters of the MOS tube and simulate its working environment in actual applications;

[0148] The acquisition module uses an optical micro-interferometer to capture the charge distribution changes and thermal effects generated by the MOS tube during the switching process, obtain the initial time domain response data, and convert it into frequency domain data;

[0149] The analysis module is used to comprehensively process the time domain and frequency domain response data, evaluate the response rate of the MOS tube under different frequencies and working conditions, and generate a dynamic response rate report;

[0150] The prediction model module trains and optimizes the response rate prediction model based on historical data and the generated dynamic response rate report to predict the response rate of the future MOS tube;

[0151] The adjustment module is used to compare the actual test results with the predicted results, and dynamically adjust the working parameters of the MOS tube through the error feedback mechanism mean square error MSE to optimize its response rate.

[0152] This embodiment also provides a computer device, which is suitable for the case of a response rate testing method for a power MOS tube, comprising: a memory and a processor; the memory is used to store computer executable instructions, and the processor is used to execute computer executable instructions to implement the response rate testing method for a power MOS tube as proposed in the above embodiment.

[0153] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covered on the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0154] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, the response rate test method of a power MOS tube proposed in the above embodiment is implemented; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, referred to as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, referred to as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, referred to as EPROM), programmable read-only memory (Programmable Red-Only Memory, referred to as PROM), read-only memory (Read-Only Memory, referred to as ROM), magnetic storage, flash memory, magnetic disk or optical disk.

[0155] In summary, the present invention provides a power MOS tube response rate testing method based on an optical microinterferometer, which can effectively capture the dynamic changes of charge distribution and thermal effects of the MOS tube during the switching process, and provide more accurate time domain response data. At the same time, the time domain data is converted into frequency domain data through fast Fourier transform FFT, and the response rate of the MOS tube at different frequencies is comprehensively evaluated. This method can not only generate a dynamic response rate report, but also dynamically adjust the working parameters of the MOS tube by training and optimizing the response rate prediction model through historical data, combining the comparison between the actual test results and the predicted results, thereby realizing the optimization and improvement of the response rate of the MOS tube. Our invented method has high accuracy and dynamic adaptability, especially under complex working conditions of high frequency and high voltage, and can significantly improve the response rate evaluation and optimization effect of the power MOS tube.

[0156] Example 2, referring to Table 1, is the second example of the present invention. To further verify the technical solution of the present invention, experimental simulation data of a response rate test method for a power MOS tube are provided.

[0157] This experiment aims to verify that the time domain response signal is converted into a frequency domain signal by fast Fourier transform (FFT) to analyze the frequency response characteristics of the power MOS tube under high-frequency switching conditions. Two groups of power MOS tubes are selected in the experiment, and the FFT method of the present invention and the prior art are used for comparative tests. The test objects are three N-type MOSFET devices of different specifications, which are marked as sample A, sample B and sample C, respectively. They are mainly used in power conversion modules of electric vehicles and high-frequency switching power supplies, with specifications of 600V and 20A;

[0158] The preparation for the experiment includes the preliminary setting of the test equipment and power MOS tube. First, the power supply voltage is set to 450V, the gate drive voltage is set to 15V, and the test frequency range is from 10kHz to 1MHz. The experiment focuses on the following parameters: switching frequency, response rate, power loss, charge distribution stabilization time, thermal stress and phase shift. These parameters are used to evaluate the performance of MOS tubes under high frequency conditions;

[0159] The experiment is divided into two parts. The first part is a control experiment using the existing technology. It mainly measures the time domain current and voltage waveforms of the MOS tube through an oscilloscope, and observes the current change amplitude and phase at different frequencies. Since the existing technology can only capture time domain signals and fails to obtain frequency domain signals, it is impossible to conduct in-depth frequency response characteristic analysis. The second part is a test based on the method of the present invention. The experiment uses an optical micro-interferometer and an electrical signal acquisition device, combined with a fast Fourier transform FFT algorithm, to convert the time domain response data into Convert to frequency domain signal , by analyzing the frequency domain signal, the amplitudes are calculated respectively. and Phase , thus obtaining the dynamic response characteristics of the MOS tube at different frequencies;

[0160] During the experiment, the optical micro-interferometer is used to accurately detect the charge distribution and thermal effect changes of the MOS tube during high-frequency switching. Through the changes in the interference fringes, the real-time data of the tiny deformation of the MOS tube surface is obtained, and these data are combined with the electrical response of the MOS tube for comprehensive analysis. The experiment uses the formula Calculate the change in charge distribution and use the formula The physical length change caused by thermal effects is evaluated, and then the time domain signal is transformed into Convert to frequency domain signal , and analyze the amplitude and phase of the frequency domain signal;

[0161] The details are shown in Table 1 below:

[0162] Table 1 Test record

[0163]

[0164] Through the analysis of experimental data, it can be clearly seen that the present invention has shown significant advantages in multiple key parameters. First, in terms of response rate improvement, the sample using the method of the present invention shows obvious advantages. For example, at a frequency of 200kHz, the response rate of sample B of the present invention is improved by 25%, while the prior art cannot accurately capture frequency domain data and fails to effectively improve the response rate. This shows that the present invention can capture and optimize the dynamic response characteristics of the MOS tube in the high-frequency switch through the fast Fourier transform FFT, thereby making the response of the switching process faster;

[0165] Secondly, the comparison of charge distribution stabilization time also fully proves the superiority of the present invention. At a frequency of 500kHz, the charge distribution stabilization time of the prior art is 2.8μs, while that of sample C of the present invention is only 1.5μs, which is reduced by nearly 46%. This shows that by combining the optical micro-interferometer with FFT, the present invention can stabilize the charge distribution faster, reduce the transition time, and improve the working efficiency of the MOS tube.

[0166] Thermal effect is an important factor affecting the performance of power MOS tubes in high-frequency switches, but the existing technology has limited monitoring capabilities in this regard and cannot provide relevant data. The present invention can accurately detect the physical expansion caused by thermal effect through an optical microinterferometer. The data show that at a frequency of 500kHz, the surface deformation of sample C is only 7.8nm, which is much lower than the traditional thermal effect expansion value under low-frequency conditions. This shows that the present invention can effectively control the thermal effect under high-frequency conditions, reduce power consumption and improve the long-term stability of MOS tubes.

[0167] The comparison of maximum power loss further proves the advantages of the present invention. Under the condition of 500kHz frequency, the power loss of sample C of the prior art reaches 6.0W, while that of sample C of the present invention is only 2.8W, which is reduced by more than 53%. This result shows that through the precise analysis of frequency domain characteristics, the present invention can optimize the energy loss in the switching process and significantly improve the power efficiency.

[0168] Signal phase shift is also a key parameter for measuring the performance of MOS tubes. From the table data, it can be seen that the phase shift of the prior art is 8.0° at a frequency of 500kHz, while the phase shift of sample C of the present invention is only 1.9°, which shows that the present invention can significantly reduce the delay and distortion in the signal transmission process and maintain higher signal synchronization and stability under high-frequency switching conditions;

[0169] Finally, the comparison of the interference fringe stabilization time also shows the control ability of the present invention on the thermal effect under high frequency conditions. Through the measurement of the micro-interferometer, the interference fringe stabilization time of the sample C of the present invention at a frequency of 500kHz is 0.20ms. This result shows that the thermal stress of the MOS tube is quickly stabilized, thereby reducing the performance fluctuation caused by the thermal effect.

[0170] In summary, through detailed analysis of the table data, it can be seen that the present invention has significant advantages in high-frequency response rate, charge distribution stability, thermal effect control, power loss reduction and signal phase shift. Compared with the prior art, the present invention can more comprehensively analyze and optimize the dynamic performance of the MOS tube through the combination of FFT and optical micro-interferometer, especially under high-frequency working conditions, and its innovation and practicality have been fully verified;

[0171] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for testing the response rate of a power MOS tube, characterized in that: include: Set the working parameters of the MOS tube to simulate its working environment in actual applications; The charge distribution and thermal effects generated by the MOS tube during the switching process are captured by an optical micro-interferometer to obtain the initial time domain response data; Convert the time domain response data into frequency domain data to obtain the frequency response characteristics; Combine time domain response data and frequency domain data to comprehensively evaluate the response rate of MOS tubes at different frequencies and working conditions, and generate a dynamic response rate report; Build a response rate prediction model based on the dynamic response rate report, train and optimize the response rate prediction model through historical data, input the working parameters of the MOS tube into the response rate prediction model, and predict the response rate of the MOS tube in the future; Compare the actual test results with the predicted results, and dynamically adjust the operating parameters of the MOS tube according to the error.

2. The method for testing the response rate of a power MOS tube according to claim 1, wherein: The setting of the working parameters of the MOS tube simulates its working environment in actual application, including: Set the voltage, current and frequency of the MOS tube; Collect the gate-source voltage and drain-source voltage of the MOS tube and calculate the working current of the MOS tube. The expression is: ； in, is the drain current, is the drain-source voltage, is the on-resistance of the MOS tube, represents the base of natural logarithms, is the gate-source voltage, Indicates the threshold voltage of the MOS tube; The switching frequency of the MOS tube is calculated by the time between the off state and the on state of the MOS tube. The expression is: ； in, represents the switching frequency, Indicates the time it takes for the MOS tube to switch from the off state to the on state. Indicates the time it takes for the MOS tube to switch from the on state to the off state.

3. The method for testing the response rate of a power MOS tube according to claim 2, wherein: The optical micro-interferometer is used to capture the charge distribution and thermal effect generated by the MOS tube during the switching process to obtain the initial time domain response data. The specific steps are: An optical micro-interferometer is used to illuminate the area from the gate to the source of the MOS tube to measure the deformation and thermal effects of the area; After the working conditions of the MOS tube are set, the optical micro-interferometer is started to capture the changes in charge distribution and thermal effects generated by the MOS tube during the switching process; Calculate the charge distribution change of the MOS tube during the switching process, the expression is: ； in, Indicates at time The charge distribution of represents the maximum accumulated charge, represents the time constant of the change of charge distribution; Calculate the change of thermal effect of MOS tube during switching process, the expression is: ； in, Indicates at time The physical length change, Indicates the initial length of the MOS tube surface before work. It indicates the rate of change of the linear dimension of a material under unit temperature change. Indicates time The temperature change; The response rate of the MOS tube is calculated based on the change in charge distribution and thermal effect. The expression is: ； in, Indicates that the MOS tube is in time The response rate, represents the weight of the charge distribution change, The weight representing the change in thermal effect.

4. The method for testing the response rate of a power MOS tube according to claim 3, wherein: The time domain response data is converted into frequency domain data to obtain the frequency response characteristics, specifically, Using Fast Fourier Transform FFT Convert to frequency domain signal ; By analyzing the frequency domain signal The frequency response characteristics are obtained, and the amplitude and phase calculation formulas of the frequency domain signal are: ； ； in, Indicates the switching frequency The signal amplitude at Indicates the switching frequency The signal phase at Represents the argument function of a complex number.

5. The method for testing the response rate of a power MOS tube according to claim 4, wherein: The time domain response data and frequency domain data are combined to comprehensively evaluate the response rate of the MOS tube under different frequencies and different working conditions, and generate a dynamic response rate report, specifically, Combine the response rate of the MOS tube, the amplitude and phase of the frequency domain signal; The response rate of the MOS tube at different frequencies and working conditions is comprehensively evaluated to obtain the comprehensive response rate, which is expressed as: ； in, Indicates time, is the comprehensive response rate, represents the weight of the time domain response data, Represents the weight of frequency domain response data; According to the overall response rate Generates a dynamic response rate report that includes a comprehensive response rate curve, frequency response characteristic analysis, and time domain response characteristic analysis.

6. The method for testing the response rate of a power MOS tube according to claim 5, wherein: The response rate prediction model is constructed according to the dynamic response rate report, and the response rate prediction model is trained and optimized through historical data. The specific steps are: Based on the generated dynamic response rate report, a response rate prediction model is constructed, the neural network model NeuralNetwork is selected as the response rate prediction model, and the multi-layer perceptron MLP is selected as the basic structure of the neural network; The gate-source voltage, drain-source voltage, switching frequency, on-resistance, time domain response rate, frequency domain amplitude and frequency domain phase of the MOS tube are used as inputs of the response rate prediction model; The multi-layer formula for predicting response rate is: ； ； ； in, , , , , , and They correspond to the weights of the gate-source voltage, drain-source voltage, switching frequency, on-resistance, time domain response rate, frequency domain amplitude and frequency domain phase of the MOS tube, respectively. represents the hyperbolic tangent function, represents the intermediate features passed from the first layer, and represents the parameters with training, and They represent the first and second nonlinear mapping functions of the response rate prediction model, represents the predicted response rate; The inputting of the working parameters of the MOS tube into the response rate prediction model to predict the response rate of the MOS tube in the future period of time refers to combining the nonlinear mapping functions of the first layer and the second layer to obtain ; Select mean square error MSE as the loss function to measure the gap between the predicted value and the true value; A large amount of historical data is extracted from the dynamic response rate report, and the data set is divided into a training set and a test set. The training set data is input into the response rate prediction model as training, and the network weights are continuously adjusted through the back propagation algorithm. In each iteration, the error between the predicted response rate and the actual response rate is calculated, and the loss function is minimized through the gradient descent method; The error between the predicted results and the actual test results is calculated using the test set.

7. The method for testing the response rate of a power MOS tube according to claim 6, wherein: The actual test results are compared with the predicted results, and the working parameters of the MOS tube are dynamically adjusted according to the error, specifically, Use optical micro-interferometer and related sensors to obtain the response rate data of MOS tube under the current working parameter conditions, and set the actual response rate collected to ; Use mean square error (MSE) as the loss function, the error between the predicted result and the actual test result (MSE); Dynamically update each parameter value according to the error value MSE, the expression is: ； in, Indicates the working parameters to be adjusted, represents the number of iterations, represents the learning rate, represents partial derivative.

8. A power MOS tube response rate test system, based on the power MOS tube response rate test method according to any one of claims 1 to 7, characterized in that: It includes setting module, collection module, analysis module, prediction model module and adjustment module. The setting module is used to set and control the working parameters of the MOS tube to simulate its working environment in actual applications; The acquisition module captures the charge distribution changes and thermal effects generated by the MOS tube during the switching process through an optical micro-interferometer, obtains initial time domain response data, and converts it into frequency domain data; The analysis module is used to comprehensively process the time domain and frequency domain response data, evaluate the response rate of the MOS tube under different frequencies and working conditions, and generate a dynamic response rate report; The prediction model module trains and optimizes the response rate prediction model based on historical data and the generated dynamic response rate report to predict the response rate of the future MOS tube; The adjustment module is used to compare the actual test results with the predicted results, and dynamically adjust the working parameters of the MOS tube through the error feedback mechanism mean square error MSE to optimize its response rate.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the power MOS tube response rate testing method described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the power MOS tube response rate testing method according to any one of claims 1 to 7 are implemented.

Images