MOS field effect tube threshold voltage adjustment method, device, equipment and storage medium

Abstract

The invention relates to the technical field of voltage regulation, and discloses a MOS field effect tube threshold voltage regulation method, device, equipment and storage medium, wherein the method comprises: scanning multiple groups of gate voltages and drain voltages of the MOS field effect tube based on initial structural parameters to obtain a multi-dimensional voltage-current characteristic data matrix; performing feature extraction and fitting to obtain an initial threshold voltage function; calculating an actual threshold voltage value and performing comparison to obtain a threshold voltage deviation; designing ion implantation parameters to perform structural parameter regulation to obtain the regulated structural parameters; performing multi-temperature point threshold voltage simulation to obtain a temperature-threshold voltage response curve; constructing an adaptive PID controller and performing dynamic compensation of the threshold voltage, and outputting a real-time threshold voltage optimization strategy, wherein the method can comprehensively consider multiple factors such as device structural parameters, process fluctuations, temperature influence, etc., to achieve accurate control and dynamic compensation of the threshold voltage.

Description

Technical Field

[0001] The present invention relates to the field of voltage regulation technology, and in particular to a method, device, equipment and storage medium for regulating the threshold voltage of a MOS field effect transistor. Background Art

[0002] As the core device of modern integrated circuits, precise control of the threshold voltage of MOS field-effect transistors is crucial to improving chip performance and reliability. As the integrated circuit process continues to shrink, the fluctuation and instability of the threshold voltage have become increasingly prominent, posing a huge challenge to chip design and manufacturing. Traditional threshold voltage adjustment methods mainly rely on process methods such as ion implantation and annealing, but these methods often have problems such as insufficient accuracy and poor adaptability, and are difficult to meet the strict requirements of advanced process nodes.

[0003] In addition, with the diversification of application environments, MOS field effect transistors are also facing the problem of threshold voltage drift caused by factors such as temperature changes and working state switching. This drift not only affects the normal operation of the device, but may also cause circuit function failure or reduced reliability. Most existing compensation technologies use fixed parameter methods, which are difficult to cope with complex and changeable actual working environments, especially under wide temperature ranges and high dynamic working conditions, and their effects are often unsatisfactory. Summary of the invention

[0004] The present invention provides a MOS field effect tube threshold voltage adjustment method, device, equipment and storage medium. The present invention can comprehensively consider multiple factors such as device structure parameters, process fluctuations, temperature effects, etc. to achieve accurate control and dynamic compensation of threshold voltage.

[0005] In a first aspect, the present invention provides a method for adjusting a threshold voltage of a MOS field effect transistor, the method comprising:

[0006] Based on the initial structural parameters, multiple groups of gate voltage and drain voltage are scanned for the MOS field effect tube to obtain a multi-dimensional voltage-current characteristic data matrix;

[0007] Inputting the multidimensional voltage-current characteristic data matrix into a support vector regression model to perform feature extraction and fitting to obtain an initial threshold voltage function;

[0008] Calculating an actual threshold voltage value based on the initial threshold voltage function, and comparing it with a preset target threshold voltage value to obtain a threshold voltage deviation;

[0009] According to the threshold voltage deviation, ion implantation parameters are designed using a particle swarm optimization algorithm, and structural parameters of the MOS field effect transistor are regulated to obtain regulated structural parameters;

[0010] Inputting the regulated structural parameters into a temperature sensitive circuit model, performing a multi-temperature threshold voltage simulation, and obtaining a temperature-threshold voltage response curve;

[0011] Based on the temperature-threshold voltage response curve, an adaptive PID controller is constructed, and the threshold voltage of the MOS field effect tube is dynamically compensated to output a real-time threshold voltage optimization strategy.

[0012] In a second aspect, the present invention provides a MOS field effect transistor threshold voltage adjustment device, the MOS field effect transistor threshold voltage adjustment device comprising:

[0013] A scanning module is used to scan multiple groups of gate voltages and drain voltages of the MOS field effect transistor based on the initial structural parameters to obtain a multi-dimensional voltage-current characteristic data matrix;

[0014] A fitting module, used for inputting the multi-dimensional voltage-current characteristic data matrix into a support vector regression model, performing feature extraction and fitting, and obtaining an initial threshold voltage function;

[0015] A comparison module, used for calculating an actual threshold voltage value based on the initial threshold voltage function, and comparing it with a preset target threshold voltage value to obtain a threshold voltage deviation;

[0016] A control module, used to design ion implantation parameters according to the threshold voltage deviation using a particle swarm optimization algorithm, and to control the structural parameters of the MOS field effect transistor to obtain the controlled structural parameters;

[0017] A simulation module, used for inputting the regulated structural parameters into a temperature sensitive circuit model, performing multi-temperature point threshold voltage simulation, and obtaining a temperature-threshold voltage response curve;

[0018] The output module is used to construct an adaptive PID controller based on the temperature-threshold voltage response curve, dynamically compensate the threshold voltage of the MOS field effect tube, and output a real-time threshold voltage optimization strategy.

[0019] The third aspect of the present invention provides a MOS field effect transistor threshold voltage adjustment device, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor calls the instructions in the memory so that the MOS field effect transistor threshold voltage adjustment device executes the above-mentioned MOS field effect transistor threshold voltage adjustment method.

[0020] A fourth aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the above-mentioned MOS field effect transistor threshold voltage adjustment method.

[0021] In the technical solution provided by the present invention, by combining a multidimensional voltage-current characteristic data matrix and a support vector regression model, the present method can more accurately extract and fit the threshold voltage characteristics of a MOS field effect tube, significantly improving the accuracy of threshold voltage regulation. The particle swarm optimization algorithm is used to design ion implantation parameters, and combined with a temperature-sensitive circuit model, the present method can automatically adjust the adjustment strategy according to the actual working environment, greatly enhancing the adaptive ability of the system. By constructing a temperature-threshold voltage response curve and designing an adaptive PID controller based on the curve, the present method realizes real-time and accurate temperature compensation for the threshold voltage, effectively solving the problem of temperature drift. The PID controller parameters are updated by real-time monitoring and recursive least squares method, so that the present method can quickly respond and adjust when the working state of the MOS field effect tube changes, improving the dynamic adjustment ability. By comprehensively considering multiple influencing factors and performing comprehensive parameter optimization, the present method significantly improves the stability and reliability of the MOS field effect tube under various working conditions. Combined with initial structural parameter modeling and ion implantation parameter optimization, the present invention has adaptive ability and can automatically adjust the compensation strategy according to the real-time working state to ensure that the MOS field effect tube can maintain stable and reliable performance under various conditions.

[0022] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention are realized and obtained by the structures particularly pointed out in the description, claims and drawings.

[0023] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 A schematic diagram of an embodiment of a method for adjusting a threshold voltage of a MOS field effect transistor according to an embodiment of the present invention;

[0025] Figure 2 A schematic diagram of an embodiment of a MOS field effect transistor threshold voltage adjustment device in an embodiment of the present invention;

[0026] Figure 3 It is a schematic diagram of an embodiment of a MOS field effect transistor threshold voltage adjustment device in an embodiment of the present invention. DETAILED DESCRIPTION

[0027] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0028] The terms "including" and "having" and any variations thereof mentioned in the embodiments of the present invention are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device end including a series of steps or units is not limited to the listed steps or units, but may optionally include other steps or units that are not listed, or may optionally include other steps or units that are inherent to these processes, methods, products or device ends.

[0029] To facilitate understanding of this embodiment, a MOS field effect transistor threshold voltage adjustment method disclosed in an embodiment of the present invention is first described in detail. Figure 1 As shown, the method comprises the following steps:

[0030] 101. Scan multiple groups of gate voltages and drain voltages of the MOS field effect transistor based on the initial structural parameters to obtain a multi-dimensional voltage-current characteristic data matrix;

[0031] It is understandable that the execution subject of the present invention may be a MOS field effect transistor threshold voltage adjustment device, or a terminal or a server, which is not limited here. The embodiment of the present invention is described by taking a server as the execution subject as an example.

[0032] Specifically, the initial structural parameters of the MOS field effect tube are parameterized and modeled, and a device model function is established, which can reflect the physical characteristics and performance of the device. By modeling the structural parameters, the performance of the MOS tube under different operating conditions can be effectively predicted. The scanning matrix of the gate voltage and the drain voltage is generated based on the device model function. A voltage scanning sequence is designed to ensure that the selected gate voltage and drain voltage cover the working range of the MOS field effect tube so as to obtain a comprehensive voltage-current characteristic. The voltage scanning sequence is input into a high-precision digital-to-analog converter for analog voltage signal conversion to obtain the actual scanning voltage. The actual scanning voltage is applied to the MOS field effect tube, and the corresponding current response is collected in real time through a high-speed sampling circuit to obtain the original current data. The original current data is digitally filtered to remove the existing noise and interference to obtain cleaner and more reliable current data. The filtered current data is paired with the corresponding actual scanning voltage to construct the voltage-current characteristic data point. Each data point represents the current response of the MOS field effect tube under specific voltage conditions. The voltage-current characteristic data point is subjected to principal component feature dimensionality reduction to effectively reduce the dimension of the data while retaining as much original information as possible. This process helps eliminate redundant information, making subsequent modeling and analysis more efficient. Through the characteristic data obtained after dimensionality reduction, a multidimensional interpolation function is constructed to obtain a continuous voltage-current characteristic surface, which reflects the behavior mode of MOS field effect tubes under different voltage conditions. The continuous voltage-current characteristic surface is meshed and the characteristic surface is converted into discrete characteristic data, which are then organized into a matrix form. Normalizing the matrix data helps to improve the consistency and comparability of the data, and finally obtains a multidimensional voltage-current characteristic data matrix.

[0033] 102. Input the multidimensional voltage-current characteristic data matrix into a support vector regression model to perform feature extraction and fitting to obtain an initial threshold voltage function;

[0034] Specifically, the kernel matrix calculation is performed on the multidimensional voltage-current characteristic data matrix, and the radial basis function is selected as the kernel function, which can effectively map the input data to the high-dimensional feature space and enhance the model's fitting ability for complex data structures. Through feature space mapping, the intrinsic relationship of the data is highlighted, so that the subsequent support vector regression model can better fit the data and obtain preliminary fitting results. The preliminary fitting results are cross-validated to evaluate the model performance. By dividing the data set into multiple subsets and training the model on different training and validation sets, the model performance evaluation indicators, including mean square error, determination coefficient, etc., are obtained to reflect the fitting effect and prediction ability of the model. Based on the evaluation indicators, the hyperparameters of the support vector regression model, such as regularization parameters and kernel function parameters, are adjusted to obtain optimized model parameters, improve the generalization ability of the model, and make its performance on new data more stable. Using the optimized model parameters, the multidimensional voltage-current characteristic data matrix is ​​refitted to generate more accurate fitting results, reflecting the actual behavior of the MOS field effect tube under different voltage conditions. After the fitting is completed, the de-normalization process is performed to map the fitting results from the feature space back to the original scale to obtain the initial threshold voltage data points. Based on the initial threshold voltage data points, a piecewise linear interpolation function is constructed to form a continuous threshold voltage curve. By interpolating discrete data points, a smooth threshold voltage curve is generated to reflect the trend of threshold voltage changing with gate voltage and drain voltage. The continuous threshold voltage curve is numerically differentiated, and the first-order derivative of the threshold voltage curve is calculated to reveal the threshold voltage change rate, that is, the sensitivity of the threshold voltage to input voltage changes. Combining the threshold voltage curve and the threshold voltage change rate, an initial threshold voltage function is constructed to describe the threshold voltage state under specific conditions and provide a theoretical basis for dynamic adjustment of the threshold voltage.

[0035] 103. Calculate the actual threshold voltage value based on the initial threshold voltage function, and compare it with the preset target threshold voltage value to obtain the threshold voltage deviation;

[0036] Specifically, the current operating point parameter of the MOS field effect tube is input into the initial threshold voltage function to generate a preliminary threshold voltage calculation value. Since the threshold voltage is affected by the ambient temperature, the preliminary threshold voltage calculation value is temperature compensated, and the calculation result is corrected according to the current ambient temperature to obtain the threshold voltage value after temperature correction. The temperature-corrected threshold voltage value is compared with the preset ideal threshold voltage model, and the deviation coefficient is calculated to obtain the correction coefficient. Based on the correction coefficient, the temperature-corrected threshold voltage value is calibrated to more accurately reflect the performance of the MOS field effect tube under the ideal working state. The difference between the calibrated actual threshold voltage value and the preset target threshold voltage value is calculated to generate an initial threshold voltage deviation. The mean and standard deviation of the initial threshold voltage deviation are calculated to describe the distribution characteristics of the threshold voltage deviation. By evaluating the distribution of the deviation, potential outliers are identified. In the process of outlier detection, the threshold voltage deviation distribution is analyzed to obtain a valid threshold voltage deviation set. The set can exclude extreme values ​​and focus on the deviation data that actually affects the circuit performance. Weighted averaging of the valid threshold voltage deviation set helps to eliminate the influence of individual outliers on the overall calculation results and obtain a more representative weighted average deviation value. Compare the weighted average deviation value with the preset allowable deviation range to determine whether it exceeds the set threshold, reveal the degree of deviation, and obtain the determination result of the degree of deviation. Based on the determination result, calculate the threshold voltage deviation to formulate corresponding adjustment strategies to ensure that the MOS field effect tube can work stably under various environmental conditions and improve its reliability and adaptability.

[0037] 104. According to the threshold voltage deviation, the particle swarm optimization algorithm is used to design ion implantation parameters, and the structural parameters of the MOS field effect tube are regulated to obtain the regulated structural parameters;

[0038] Specifically, the objective function of ion implantation is set based on the threshold voltage deviation. The objective function converts the deviation into the optimization target of implantation dose and energy, forming a clear parameter optimization model. The optimization model is initialized to generate a particle swarm including implantation dose, implantation energy and implantation angle to form an initial particle swarm. The initial particle swarm is used as the basis of the optimization algorithm to find the optimal solution through continuous iteration. After the initial particle swarm is input into the ion implantation simulator, the implantation process is simulated to obtain the impurity distribution curve corresponding to each particle, and the fitness value of each particle is calculated based on these curves. The effectiveness of the particle in regulating the threshold voltage is evaluated to obtain the particle fitness ranking. According to the particle fitness ranking result, the optimal position of the particle swarm is updated, and the speed and position of the particle are adjusted at the same time to obtain the updated particle swarm. In this process, the optimization ability of the particle swarm is continuously enhanced. By introducing fitness information, the particle moves towards the direction of the optimal solution. The updated particle swarm will repeat the implantation simulation and fitness calculation until the preset number of iterations is reached. In each iteration, the speed and position of the particle are continuously adjusted, and the ion implantation parameters are continuously optimized to obtain the optimal implantation parameters. The optimal ion implantation parameters are used to perform actual ion implantation on the MOS field effect transistor. This process precisely controls the dose, energy, and angle of the implantation equipment to ensure that the implanted structure meets the design requirements. After the ion implantation is completed, the implanted structure is subjected to rapid thermal annealing to activate the implanted impurity atoms and repair lattice defects. After the thermal treatment is completed, the key electrical parameters of the thermally treated structure, such as threshold voltage, subthreshold swing, and leakage current, are measured to provide detailed information about the structural performance. These electrical parameters constitute a structural parameter data set, based on which the device model is extracted to obtain the regulated structural parameters.

[0039] 105. Input the adjusted structural parameters into the temperature sensitive circuit model, perform multi-temperature point threshold voltage simulation, and obtain a temperature-threshold voltage response curve;

[0040] Specifically, the regulated structural parameters are input into the temperature-sensitive circuit model to obtain a temperature-dependent device model, which can reflect the electrical performance of the MOS field effect tube at different temperatures. Parameters of the temperature-dependent device model are extracted to obtain a set of temperature-sensitive parameter sets. Multiple temperature points are set based on the temperature-sensitive parameter set to form a temperature point set. These temperature points should cover the working range of the MOS field effect tube to ensure a comprehensive evaluation of its performance. For each temperature point in the set, circuit simulation is performed to generate a corresponding voltage-current characteristic curve for each temperature point to show the current response and voltage characteristics of the device at a specific temperature. From the voltage-current characteristic curve, the threshold voltage value of each temperature point is extracted to form a discrete temperature-threshold voltage data point. Data processing is performed on the discrete temperature-threshold voltage data points to obtain a smoothed data set. Based on the smoothed data set, the relationship between temperature and threshold voltage is fitted to obtain a response curve equation, which describes how the threshold voltage changes under different temperature conditions. After the fitting is completed, the response curve equation is subjected to error analysis to evaluate the accuracy of the model and determine the error range of the response curve. Based on the response curve equation and its error range, a final temperature-threshold voltage response curve is generated.

[0041] 106. Based on the temperature-threshold voltage response curve, an adaptive PID controller is constructed, and the threshold voltage of the MOS field effect tube is dynamically compensated to output a real-time threshold voltage optimization strategy.

[0042] Specifically, the temperature-threshold voltage response curve is curve-fitted to generate a functional relationship between the threshold voltage and the temperature. Based on the functional relationship, the temperature coefficient is calculated to obtain the sensitivity parameter of the threshold voltage changing with the temperature, so as to ensure that the controller can appropriately adjust the threshold voltage according to the temperature change. The sensitivity parameter is input into the PID controller to preliminarily set the proportional, integral and differential parameters of the PID controller, generate a set of initial PID parameter groups, and construct a feedback control loop based on the parameter group. The loop can calculate the threshold voltage deviation by comparing the threshold voltage measured in real time with the target threshold voltage, reflecting the difference between the actual threshold voltage and the target value. According to the calculated threshold voltage deviation, the PID controller outputs a signal to obtain a preliminary threshold voltage compensation value. The preliminary compensation value is subjected to a limiting process, and the maximum compensation range is set to avoid the negative impact of excessive adjustment on the system. After the limiting process, the effective compensation signal obtained is used to further adjust the gate voltage of the MOS field effect tube. After the effective compensation signal is converted into a gate voltage adjustment amount, the compensated threshold voltage is achieved by adjusting the gate voltage of the MOS field effect tube. The compensated threshold voltage is monitored in real time, the compensation effect is recorded, and the threshold voltage adjustment history data is formed. Based on the historical data of threshold voltage adjustment, the recursive least squares method is used to update the parameters of the PID controller. By analyzing historical data, this method can optimize the parameters of the PID controller to make it more adaptable and accurate in actual work. The optimized PID parameters are applied to the controller, combined with the current temperature information, and a real-time threshold voltage optimization strategy is output to ensure that the threshold voltage of the MOS field effect tube can remain stable under various working conditions, thereby improving the performance and reliability of the device. By continuously looping this process, it self-adjusts in a dynamic environment to adapt to different operating conditions and achieves efficient threshold voltage control.

[0043] In the embodiment of the present invention, by combining the multidimensional voltage-current characteristic data matrix and the support vector regression model, the method can more accurately extract and fit the threshold voltage characteristics of the MOS field effect tube, and significantly improve the accuracy of the threshold voltage adjustment. The particle swarm optimization algorithm is used to design the ion implantation parameters, and combined with the temperature sensitive circuit model, the method can automatically adjust the adjustment strategy according to the actual working environment, greatly enhancing the adaptive ability of the system. By constructing a temperature-threshold voltage response curve and designing an adaptive PID controller based on it, the method realizes real-time and accurate temperature compensation for the threshold voltage, and effectively solves the temperature drift problem. The PID controller parameters are updated by real-time monitoring and recursive least squares method, so that the method can quickly respond and adjust when the working state of the MOS field effect tube changes, and improves the dynamic adjustment ability. By comprehensively considering multiple influencing factors and performing comprehensive parameter optimization, the method significantly improves the stability and reliability of the MOS field effect tube under various working conditions. Combined with the initial structure parameter modeling and ion implantation parameter optimization, the present invention has adaptive ability, and can automatically adjust the compensation strategy according to the real-time working state to ensure that the MOS field effect tube can maintain stable and reliable performance under various conditions.

[0044] In a specific embodiment, the process of executing step 101 may specifically include the following steps:

[0045] Perform parameterized modeling on the initial structural parameters of the MOS field effect tube to obtain a device model function, and generate a scanning matrix of the gate voltage and the drain voltage based on the device model function to obtain a voltage scanning sequence;

[0046] The voltage scanning sequence is input into a high-precision digital-to-analog converter for analog voltage signal conversion to obtain the actual scanning voltage, and the actual scanning voltage is applied to the MOS field effect transistor, and the corresponding current response is collected using a high-speed sampling circuit to obtain the original current data;

[0047] Perform digital filtering on the original current data to obtain filtered current data, and pair the filtered current data with the corresponding actual scan voltage to construct voltage-current characteristic data points;

[0048] The voltage-current characteristic data points are subjected to principal component feature dimensionality reduction to obtain feature data after dimensionality reduction, and a multidimensional interpolation function is constructed based on the feature data after dimensionality reduction to obtain a continuous voltage-current characteristic surface;

[0049] The continuous voltage-current characteristic surface is meshed to obtain discretized characteristic data, the discretized characteristic data is organized into a matrix form, and normalized to obtain a multi-dimensional voltage-current characteristic data matrix.

[0050] Specifically, the initial structural parameters of the MOS field effect tube are parametrically modeled and analyzed, including gate aspect ratio, doping concentration, oxide layer thickness, etc. These parameters are used to construct a device model function that describes the electrical characteristics of the MOS tube. A scanning matrix of gate voltage and drain voltage is generated based on the device model function. The matrix uses a multidimensional array to represent different voltage combinations. For example, the gate voltage and drain voltage Set as a set of multiple discrete values ​​to form a The matrix of is the number of gate voltage points, is the number of drain voltage points, and the voltage scanning sequence is obtained. The voltage scanning sequence is input into a high-precision digital-to-analog converter for analog voltage signal conversion to obtain the actual scanning voltage signal. After applying the actual scanning voltage to the MOS field effect transistor, the corresponding current response is collected using a high-speed sampling circuit to obtain the original current data. Assume that the collected current data is ,but Indicates the current value, It is a time series. The original current data is digitally filtered to effectively remove noise and interference and obtain the filtered current data. . Pair the filtered current data with the corresponding actual scan voltage to construct the voltage-current characteristic data points. Perform principal component feature dimensionality reduction on the voltage-current characteristic data points to extract the main features in the data to reduce the complexity of the data. In this process, the principal component analysis method is used to transform the high-dimensional feature space into a low-dimensional feature space to obtain the reduced-dimensional feature data. , expressed by the following formula:

[0051] ;

[0052] in, is the feature data after dimensionality reduction, For the The weights of the principal components, is the original feature data, is the number of principal components selected. Based on the feature data after dimensionality reduction, a multidimensional interpolation function is constructed, which is used to generate a continuous voltage-current characteristic surface. The continuous voltage-current characteristic surface is meshed and discretized into multiple small blocks. Each grid block corresponds to a specific voltage-current combination, and the discretized characteristic data is obtained. The discretized characteristic data is organized into a matrix form. The characteristic data matrix is ​​normalized to ensure that the data is within the same dimensional range, avoid calculation errors caused by inconsistent dimensions, and obtain a normalized multidimensional voltage-current characteristic data matrix.

[0053] In a specific embodiment, the process of executing step 102 may specifically include the following steps:

[0054] Perform kernel matrix calculation on the multi-dimensional voltage-current characteristic data matrix, select radial basis function as kernel function, obtain feature space mapping, input feature space mapping into support vector regression model, perform data fitting, and obtain preliminary fitting results;

[0055] The preliminary fitting results are cross-validated to obtain the model performance evaluation index, and the hyperparameters of the support vector regression model are adjusted based on the model performance evaluation index to obtain the optimized model parameters;

[0056] The multi-dimensional voltage-current characteristic data matrix is ​​refitted using the optimized model parameters to obtain an updated fitting result, and the updated fitting result is de-standardized to map the data back to the original scale to obtain an initial threshold voltage data point;

[0057] A piecewise linear interpolation function is constructed based on the initial threshold voltage data points to obtain a continuous threshold voltage curve, and the continuous threshold voltage curve is numerically differentiated to calculate the first-order derivative of the curve to obtain the threshold voltage change rate. The initial threshold voltage function is constructed by combining the threshold voltage curve and the threshold voltage change rate.

[0058] Specifically, a kernel matrix is ​​calculated for the multidimensional voltage-current characteristic data matrix, and a radial basis function is selected as the kernel function, where the form of the radial basis function is:

[0059] ;

[0060] in, Represents the input data point and The kernel function value between is a hyperparameter that controls the width of the kernel. Represents the Euclidean distance between two data points. By applying kernel matrix calculation to the multidimensional voltage-current characteristic data matrix, the data is mapped to a higher-dimensional feature space, making it more effective to train the support vector regression model in this space. After the feature space mapping is input into the support vector regression model, the data is fitted to obtain preliminary fitting results. In order to evaluate the performance of the model, cross-validation techniques are used. The data set is divided into several subsets, and the model is repeatedly trained and tested on these subsets to obtain model performance evaluation indicators, such as mean square error (MSE) or coefficient of determination (R²). The performance of the model in the prediction task. Based on the evaluation indicators, the hyperparameters of the support vector regression model are adjusted to obtain the optimized model parameters. The hyperparameters include (penalty parameter) and (kernel parameters), the optimization process is performed by grid search or Bayesian optimization. When the multi-dimensional voltage-current characteristic data matrix is ​​refitted using the optimized model parameters, an updated fitting result is obtained. The updated fitting result is denormalized, and the fitting data is mapped back to the original scale to obtain the initial threshold voltage data point. The denormalization process is expressed by the following formula:

[0061] ;

[0062] in, is the denormalized data. is the standardized data, is the standard deviation of the original data, and is the mean of the original data. Through this process, the threshold voltage data consistent with the original measurement is obtained. Based on the initial threshold voltage data points, a piecewise linear interpolation function is constructed. This function can generate a smooth continuous threshold voltage curve based on the known threshold voltage data points. The formula for piecewise linear interpolation is:

[0063] ;

[0064] in, is the output of the interpolation function, and The known data points are and The output value of is the input value to be solved. By constructing an interpolation function, a continuous threshold voltage curve is obtained. The continuous threshold voltage curve is numerically differentiated, and the first-order derivative of the curve is calculated to obtain the threshold voltage change rate, which represents the rate at which the threshold voltage changes with the input parameter. Combining the threshold voltage curve and the threshold voltage change rate, the initial threshold voltage function is finally constructed, so that the threshold voltage change of the MOS field effect tube can be effectively predicted under different working conditions.

[0065] In a specific embodiment, the process of executing step 103 may specifically include the following steps:

[0066] Input the current operating point parameter of the MOS field effect tube into the initial threshold voltage function to obtain a preliminary threshold voltage calculation value, and perform temperature compensation on the preliminary threshold voltage calculation value, and correct it based on the current ambient temperature to obtain a temperature-corrected threshold voltage value;

[0067] The temperature-corrected threshold voltage value is compared with a preset ideal threshold voltage model, a deviation coefficient is calculated to obtain a correction coefficient, and the temperature-corrected threshold voltage value is calibrated based on the correction coefficient to obtain a calibrated actual threshold voltage value;

[0068] The difference between the actual threshold voltage value after calibration and the preset target threshold voltage value is calculated to obtain an initial threshold voltage deviation, and the mean and standard deviation of the initial threshold voltage deviation are calculated to obtain a threshold voltage deviation distribution;

[0069] Performing outlier detection based on the threshold voltage deviation distribution to obtain a valid threshold voltage deviation set, and performing weighted averaging on the valid threshold voltage deviation set to obtain a weighted average deviation value;

[0070] The weighted average deviation value is compared with a preset allowable deviation range to determine whether it exceeds a threshold value, and a deviation degree determination result is obtained. Based on the deviation degree determination result, a threshold voltage deviation amount is calculated.

[0071] Specifically, the current operating point parameters of the MOS field effect tube are input into the initial threshold voltage function to obtain a preliminary threshold voltage calculation value. The preliminary value is affected by the ambient temperature, and the preliminary threshold voltage calculation value is temperature compensated. Assuming the current ambient temperature is , using the temperature coefficient Correction, the threshold voltage value after temperature correction Calculated by the following formula:

[0072] ;

[0073] in, At the reference temperature The threshold voltage measured under is the sensitivity of the threshold voltage to temperature. In this way, the temperature-corrected threshold voltage value can be effectively obtained. . Compare the temperature-corrected threshold voltage value with the preset ideal threshold voltage model. Calculate the deviation coefficient ,Right now:

[0074] ;

[0075] in, is the temperature corrected threshold voltage value, and It is the preset ideal threshold voltage model. The calculated deviation coefficient is used to generate the correction coefficient, which is used to calibrate the threshold voltage value after temperature correction. Set to a constant, for example:

[0076] ;

[0077] Based on the correction coefficient, the temperature-corrected threshold voltage value is calibrated to obtain the calibrated actual threshold voltage value.

[0078] ;

[0079] The actual threshold voltage value after calibration is calculated by the difference between the preset target threshold voltage value to obtain the initial threshold voltage deviation. The deviation is expressed as:

[0080] ;

[0081] in, is the preset target threshold voltage value. The mean and standard deviation of the initial threshold voltage deviation are calculated to obtain the threshold voltage deviation distribution. The mean of the calculated deviation and standard deviation

[0082] ;

[0083] ;

[0084] in, is the sample size, is the initial threshold voltage deviation of each sample. By calculating and analyzing the deviation distribution of the threshold voltage, it helps to determine whether there are outliers. Outlier detection is performed based on the threshold voltage deviation distribution to obtain a valid threshold voltage deviation set. Outlier detection is achieved by setting a threshold, for example, using the mean plus or minus two standard deviations to define the range of valid data. In this process, data that deviates from the normal range can be excluded, thereby ensuring the accuracy of subsequent processing. Weighted averaging is performed on the valid threshold voltage deviation set to obtain a weighted average deviation value , calculated by the following formula:

[0085] ;

[0086] in, is the weight of each bias, is the number of valid data. This method can more accurately reflect the current threshold voltage deviation. The weighted average deviation value is compared with the preset allowable deviation range to determine whether it exceeds the threshold value, and the deviation degree determination result is obtained to determine whether the threshold voltage needs to be further adjusted. Based on the deviation degree determination result, the threshold voltage deviation amount is calculated to achieve accurate calculation and dynamic compensation of the threshold voltage of the MOS field effect tube, ensuring stability and reliability under different working environments.

[0087] In a specific embodiment, the process of executing step 104 may specifically include the following steps:

[0088] The objective function of ion implantation is set based on the threshold voltage deviation, the deviation is converted into the optimization target of implantation dose and energy, a parameter optimization model is obtained, and the parameter optimization model is initialized to generate a particle group including implantation dose, implantation energy and implantation angle, and an initial particle group is obtained;

[0089] The initial particle group is input into the ion implantation simulator to simulate the implantation process, and the impurity distribution curve corresponding to each particle is obtained. The fitness value of each particle is calculated based on the impurity distribution curve, and its effectiveness in adjusting the threshold voltage is evaluated to obtain the particle fitness ranking.

[0090] The optimal position of the particle swarm is updated according to the particle fitness ranking, the speed and position of the particles are adjusted to obtain an updated particle swarm, and the injection simulation and fitness calculation are repeated on the updated particle swarm until a preset number of iterations is reached to obtain the optimal ion implantation parameters;

[0091] Use the optimal ion implantation parameters to perform actual ion implantation on the MOS field effect transistor, control the dose, energy and angle of the implantation equipment, obtain the implanted structure, and perform rapid thermal annealing on the implanted structure to activate the implanted impurity atoms and repair lattice defects to obtain the heat-treated structure;

[0092] The key electrical parameters of the structure after heat treatment, including threshold voltage, subthreshold swing and leakage current, are measured to obtain a structural parameter data set. The device model is extracted based on the structural parameter data set to obtain the regulated structural parameters.

[0093] Specifically, the objective function of ion implantation is set based on the threshold voltage deviation, and the deviation is converted into the optimization target of implantation dose and energy. By defining an objective function ,in Indicates the injection dose, Represents the injected energy. The objective function aims to minimize the deviation of the threshold voltage, which can be expressed as:

[0094] ;

[0095] in, is the threshold voltage deviation. By definition, the optimization goal is to adjust the threshold voltage to reach the preset target value. Based on the objective function, a parameter optimization model is constructed and initialized to generate the injection dose. , inject energy and injection angle Each particle represents a possible combination of injection parameters to form an initial particle swarm. Each particle in the particle swarm is represented by a triplet Represents the formation of particle swarm

[0096] ;

[0097] The initial particle group is input into the ion implantation simulator to simulate the implantation process. During this process, the simulator generates a corresponding impurity distribution curve based on the particle parameters to describe the spatial distribution of the injected impurities. Based on the impurity distribution curve, the fitness value of each particle is calculated. , to evaluate its effectiveness in adjusting the threshold voltage. The fitness value calculation formula is expressed as:

[0098] ;

[0099] Among them, the fitness value The higher the value, the more effective the corresponding particle parameter combination is in adjusting the threshold voltage. After obtaining the fitness values ​​of all particles, update the optimal position of the particle swarm according to the fitness ranking. Adjust the speed and position of the particles using the following update rules:

[0100] ;

[0101] ;

[0102] in, is the velocity of the particle, is the inertia weight, and is the learning factor, and is a random number, is the particle's best historical position, is the global optimal position, is the position of the current particle. Through these update rules, the updated particle swarm is obtained, and the injection simulation and fitness calculation are repeated for the particle swarm until the preset number of iterations is reached. A set of optimal ion implantation parameters is obtained. . Use the optimal ion implantation parameters to perform actual ion implantation on the MOS field effect transistor. In this process, the dose of the implantation equipment is precisely controlled. ,energy and angle , in order to achieve the design goals. After the ion implantation is completed, the implanted structure is subjected to rapid thermal annealing to activate the implanted impurity atoms and repair lattice defects, which helps to improve the electrical properties of the material and enhance the performance of the device. After the thermal treatment, the key electrical parameters of the structure are measured, including threshold voltage, subthreshold swing, and leakage current. A structural parameter data set is constructed through measurement to describe the performance of the processed MOS field effect tube. Based on the structural parameter data set, the device model is extracted to obtain the regulated structural parameters.

[0103] In a specific embodiment, the process of executing step 105 may specifically include the following steps:

[0104] Inputting the regulated structural parameters into the temperature-sensitive circuit model to obtain a temperature-dependent device model, and performing parameter extraction on the temperature-dependent device model to obtain a temperature-sensitive parameter set;

[0105] A plurality of temperature points are set based on a temperature sensitive parameter set to obtain a temperature point set, and a circuit simulation is performed on each temperature point in the temperature point set to obtain a voltage-current characteristic curve of each temperature point;

[0106] Extracting the threshold voltage value of each temperature point from the voltage-current characteristic curve to obtain discrete temperature-threshold voltage data points, and performing data processing on the discrete temperature-threshold voltage data points to obtain a smoothed data set;

[0107] The temperature-threshold voltage relationship is fitted based on the smoothed data set to obtain a response curve equation, and an error analysis is performed on the response curve equation to obtain an error range of the response curve. A temperature-threshold voltage response curve is generated based on the response curve equation and the error range.

[0108] Specifically, the regulated structural parameters are input into the temperature-sensitive circuit model to obtain a device model related to temperature. In this way, a temperature-related model describing the behavior of the MOS field effect tube at different temperatures is obtained, and parameters of the model are extracted, including the temperature sensitivity coefficient, the sensitivity of the threshold voltage to temperature change, etc. These parameters together constitute a temperature sensitive parameter set. Based on the extracted temperature sensitive parameter set, multiple temperature points are set to form a temperature point set, for example, several key temperature points such as room temperature (25°C), high temperature (85°C) and low temperature (-20°C) are selected. The temperature points are set to cover most of the environments in which the device may work and ensure the comprehensiveness and reliability of the data. On this basis, circuit simulation is performed on each temperature point to obtain a voltage-current characteristic curve to describe the working state and performance of the MOS field effect tube under different temperature conditions. After completing the circuit simulation, the threshold voltage value of each temperature point is extracted from the obtained voltage-current characteristic curve. The relationship between the characteristic curve and the current is analyzed to determine the corresponding threshold voltage. For example, the threshold voltage is set to a voltage that makes the output current reach a certain specific value. Data processing is performed on discrete temperature-threshold voltage data points to obtain a smoothed data set. The smoothing process uses moving average or other interpolation methods to eliminate noise and fluctuations in the experimental data. Based on the smoothed data set, the temperature-threshold voltage relationship is fitted to construct the response curve equation. The response curve equation is expressed as a polynomial or other suitable function form, such as:

[0109] ;

[0110] in, It is the temperature The threshold voltage under , ,and It is a coefficient determined by fitting. These coefficients reflect the specific relationship between threshold voltage and temperature. To ensure the accuracy of fitting, the response curve equation is subjected to error analysis, the difference between the fitting value and the actual data is calculated, and the error range of the response curve is obtained. Based on the response curve equation obtained by fitting and its error range, a temperature-threshold voltage response curve is generated. This curve is used to analyze the performance changes of MOS field effect tubes at different operating temperatures, providing a basis for subsequent circuit design and debugging. For example, when designing a circuit that works at high temperature, the change of threshold voltage can be foreseen according to this curve, and corresponding compensation measures can be taken to ensure that the device can work stably at different temperatures.

[0111] In a specific embodiment, the process of executing step 106 may specifically include the following steps:

[0112] Performing curve fitting on the temperature-threshold voltage response curve to obtain a functional relationship between the threshold voltage and the temperature, and calculating the temperature coefficient based on the functional relationship to obtain a sensitivity parameter of the threshold voltage changing with temperature;

[0113] The sensitivity parameter is input into the PID controller, the proportional, integral and differential parameters of the PID controller are preliminarily set, an initial PID parameter group is obtained, and a feedback control loop is constructed based on the initial PID parameter group, and the threshold voltage measured in real time is compared with the target threshold voltage to obtain the threshold voltage deviation;

[0114] The output signal of the PID controller is calculated according to the threshold voltage deviation to obtain a preliminary threshold voltage compensation value, and the preliminary threshold voltage compensation value is limited to set a maximum compensation range to obtain an effective compensation signal;

[0115] Convert the effective compensation signal into the gate voltage adjustment value, and obtain the compensated threshold voltage by adjusting the gate voltage of the MOS field effect tube. Monitor the compensated threshold voltage in real time, record the compensation effect, and obtain the threshold voltage adjustment history data.

[0116] Based on the threshold voltage adjustment historical data, the recursive least squares method is used to update the PID controller parameters to obtain the optimized PID parameters. The optimized PID parameters are applied to the controller and combined with the current temperature information to output the real-time threshold voltage optimization strategy.

[0117] Specifically, the temperature-threshold voltage response curve is curve-fitted to obtain a functional relationship between the threshold voltage and the temperature. Polynomial regression, linear regression or other nonlinear fitting techniques are used. For example, a quadratic polynomial model is set, expressed as:

[0118] ;

[0119] in, is the threshold voltage, is the temperature, and , ,and is the coefficient obtained by fitting. Through curve fitting, a clear relationship between threshold voltage and temperature is obtained, and the temperature coefficient is calculated, which is specifically expressed as the derivative of threshold voltage with respect to temperature, that is:

[0120] ;

[0121] The derivative value represents the sensitivity parameter of the threshold voltage with temperature, which determines the change of the threshold voltage under different environmental conditions. The sensitivity parameter is input into the PID controller. The proportional (P), integral (I) and differential (D) parameters of the PID controller are preliminarily set to form the initial PID parameter group. The set proportional parameter controls the immediate response of the system to the deviation, the integral parameter is responsible for eliminating the long-term steady-state error, and the differential parameter is used to reduce the overshoot of the system. Based on these parameters, a feedback control loop is constructed to compare the real-time measured threshold voltage with the preset target threshold voltage to obtain the threshold voltage deviation, which is expressed as:

[0122] ;

[0123] in, is the threshold voltage deviation, is the target threshold voltage, is the actual measured threshold voltage. , the PID controller generates an output signal to compensate for the deviation. Calculate the output value of the PID controller

[0124] ;

[0125] in, , ,and They are proportional, integral and differential gains respectively. The calculated preliminary threshold voltage compensation value is processed by limiting to set the maximum compensation range to ensure that the compensation signal does not exceed the reasonable range. The effective compensation signal is expressed as:

[0126] ;

[0127] The effective compensation signal is converted into a gate voltage adjustment value, and the compensated threshold voltage is obtained by adjusting the gate voltage of the MOS field effect tube. The threshold voltage change after compensation is monitored in real time and the compensation effect is recorded to form a historical data set for threshold voltage adjustment. Based on the historical data of threshold voltage adjustment, the PID controller parameters are updated using the recursive least squares method. Through the step-by-step analysis of historical data, the PID parameters are optimized to adapt to the current control requirements. The optimized PID parameters are expressed as:

[0128] ;

[0129] The optimized parameters are applied to the controller, and combined with the current temperature information, a real-time threshold voltage optimization strategy is output to ensure that the MOS field effect tube always maintains a good working state under different temperature environments and ensures its performance is stable and reliable.

[0130] The above describes the MOS field effect tube threshold voltage adjustment method in the embodiment of the present invention. The following describes the MOS field effect tube threshold voltage adjustment device in the embodiment of the present invention. Figure 2 In one embodiment of the present invention, a MOS field effect tube threshold voltage adjustment device includes:

[0131] A scanning module 201 is used to scan multiple groups of gate voltages and drain voltages of the MOS field effect transistor based on the initial structural parameters to obtain a multi-dimensional voltage-current characteristic data matrix;

[0132] A fitting module 202 is used to input the multi-dimensional voltage-current characteristic data matrix into a support vector regression model to perform feature extraction and fitting to obtain an initial threshold voltage function;

[0133] A comparison module 203 is used to calculate the actual threshold voltage value based on the initial threshold voltage function, and compare it with the preset target threshold voltage value to obtain a threshold voltage deviation;

[0134] The control module 204 is used to design ion implantation parameters according to the threshold voltage deviation by using a particle swarm optimization algorithm, and to control the structural parameters of the MOS field effect transistor to obtain the controlled structural parameters;

[0135] The simulation module 205 is used to input the adjusted structural parameters into the temperature sensitive circuit model, perform multi-temperature point threshold voltage simulation, and obtain a temperature-threshold voltage response curve;

[0136] The output module 206 is used to construct an adaptive PID controller based on the temperature-threshold voltage response curve, dynamically compensate the threshold voltage of the MOS field effect tube, and output a real-time threshold voltage optimization strategy.

[0137] Through the synergy of the above-mentioned components, by combining the multidimensional voltage-current characteristic data matrix and the support vector regression model, the method can more accurately extract and fit the threshold voltage characteristics of the MOS field effect tube, and significantly improve the accuracy of the threshold voltage regulation. The particle swarm optimization algorithm is used to design the ion implantation parameters, and combined with the temperature sensitive circuit model, the method can automatically adjust the adjustment strategy according to the actual working environment, greatly enhancing the adaptive ability of the system. By constructing a temperature-threshold voltage response curve and designing an adaptive PID controller based on it, the method realizes real-time and accurate temperature compensation for the threshold voltage, and effectively solves the temperature drift problem. The PID controller parameters are updated by real-time monitoring and recursive least squares method, so that the method can quickly respond and adjust when the working state of the MOS field effect tube changes, and improves the dynamic adjustment ability. By comprehensively considering multiple influencing factors and performing comprehensive parameter optimization, the method significantly improves the stability and reliability of the MOS field effect tube under various working conditions. Combined with the initial structure parameter modeling and ion implantation parameter optimization, the present invention has adaptive ability, and can automatically adjust the compensation strategy according to the real-time working state to ensure that the MOS field effect tube can maintain stable and reliable performance under various conditions.

[0138] above Figure 2 The MOS field effect transistor threshold voltage adjustment device in the embodiment of the present invention is described in detail from the perspective of modular functional entities. The MOS field effect transistor threshold voltage adjustment device in the embodiment of the present invention is described in detail from the perspective of hardware processing.

[0139] Figure 3 3 is a schematic diagram of the structure of a MOS field effect transistor threshold voltage adjustment device provided by an embodiment of the present invention. The MOS field effect transistor threshold voltage adjustment device 300 may have relatively large differences due to different configurations or performances, and may include one or more processors (central processing units, CPU) 310 (for example, one or more processors) and a memory 320, and one or more storage media 330 (for example, one or more mass storage device terminals) storing application programs 333 or data 332. Among them, the memory 320 and the storage medium 330 may be short-term storage or permanent storage. The program stored in the storage medium 330 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations in the MOS field effect transistor threshold voltage adjustment device 300. Furthermore, the processor 310 may be configured to communicate with the storage medium 330, and execute a series of instruction operations in the storage medium 330 on the MOS field effect transistor threshold voltage adjustment device 300 to implement the steps of the above-mentioned MOS field effect transistor threshold voltage adjustment method.

[0140] The MOS field effect transistor threshold voltage adjustment device 300 may also include one or more power supplies 340, one or more wired or wireless network interfaces 350, one or more input and output interfaces 360, and / or one or more operating systems 331, such as Windows Serve, Mac OS X, Unix, Linux, FreeBSD, etc. It can be understood by those skilled in the art that Figure 3 The structure of the MOS field effect transistor threshold voltage regulating device shown does not constitute a limitation on the MOS field effect transistor threshold voltage regulating device provided by the present invention, and may include more or fewer components than shown in the figure, or a combination of certain components, or a different arrangement of components.

[0141] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions are executed on a computer, the computer executes the steps of the MOS field effect transistor threshold voltage adjustment method.

[0142] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0143] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or the whole or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program code.

[0144] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for adjusting the threshold voltage of a MOS field effect tube, characterized in that: The method comprises: Based on the initial structural parameters, multiple groups of gate voltage and drain voltage are scanned for the MOS field effect tube to obtain a multi-dimensional voltage-current characteristic data matrix; Inputting the multidimensional voltage-current characteristic data matrix into a support vector regression model to perform feature extraction and fitting to obtain an initial threshold voltage function; Calculating an actual threshold voltage value based on the initial threshold voltage function, and comparing it with a preset target threshold voltage value to obtain a threshold voltage deviation; According to the threshold voltage deviation, ion implantation parameters are designed using a particle swarm optimization algorithm, and structural parameters of the MOS field effect transistor are regulated to obtain regulated structural parameters; Inputting the regulated structural parameters into a temperature sensitive circuit model, performing a multi-temperature threshold voltage simulation, and obtaining a temperature-threshold voltage response curve; Based on the temperature-threshold voltage response curve, an adaptive PID controller is constructed, and the threshold voltage of the MOS field effect tube is dynamically compensated to output a real-time threshold voltage optimization strategy.

2. The MOS field effect tube threshold voltage adjustment method according to claim 1, characterized in that: The method scans multiple groups of gate voltages and drain voltages of the MOS field effect transistor based on the initial structural parameters to obtain a multi-dimensional voltage-current characteristic data matrix, including: Performing parameterized modeling on the initial structural parameters of the MOS field effect transistor to obtain a device model function, and generating a scanning matrix of gate voltage and drain voltage based on the device model function to obtain a voltage scanning sequence; The voltage scanning sequence is input into a high-precision digital-to-analog converter for analog voltage signal conversion to obtain an actual scanning voltage, and the actual scanning voltage is applied to the MOS field effect transistor, and a high-speed sampling circuit is used to collect the corresponding current response to obtain original current data; Performing digital filtering on the raw current data to obtain filtered current data, and pairing the filtered current data with the corresponding actual scan voltage to construct voltage-current characteristic data points; Performing principal component feature dimensionality reduction on the voltage-current characteristic data points to obtain reduced-dimensional feature data, and constructing a multidimensional interpolation function based on the reduced-dimensional feature data to obtain a continuous voltage-current characteristic surface; The continuous voltage-current characteristic surface is meshed to obtain discretized characteristic data, the discretized characteristic data is organized into a matrix form, and normalized to obtain a multi-dimensional voltage-current characteristic data matrix.

3. The MOS field effect tube threshold voltage adjustment method according to claim 1, characterized in that: The multi-dimensional voltage-current characteristic data matrix is ​​input into a support vector regression model to perform feature extraction and fitting to obtain an initial threshold voltage function, including: Performing kernel matrix calculation on the multidimensional voltage-current characteristic data matrix, selecting radial basis function as kernel function, obtaining feature space mapping, and inputting the feature space mapping into a support vector regression model to perform data fitting to obtain preliminary fitting results; Cross-validating the preliminary fitting results to obtain a model performance evaluation index, and adjusting the hyperparameters of the support vector regression model based on the model performance evaluation index to obtain optimized model parameters; Refitting the multidimensional voltage-current characteristic data matrix using the optimized model parameters to obtain updated fitting results, and performing denormalization processing on the updated fitting results to map the data back to the original scale to obtain initial threshold voltage data points; A piecewise linear interpolation function is constructed based on the initial threshold voltage data points to obtain a continuous threshold voltage curve, and the continuous threshold voltage curve is numerically differentiated to calculate the first-order derivative of the curve to obtain the threshold voltage change rate, and the initial threshold voltage function is constructed by combining the threshold voltage curve and the threshold voltage change rate.

4. The MOS field effect tube threshold voltage adjustment method according to claim 1, characterized in that: The calculating the actual threshold voltage value based on the initial threshold voltage function and comparing it with the preset target threshold voltage value to obtain the threshold voltage deviation includes: Inputting the current operating point parameter of the MOS field effect tube into the initial threshold voltage function to obtain a preliminary threshold voltage calculation value, and performing temperature compensation on the preliminary threshold voltage calculation value, and correcting it based on the current ambient temperature to obtain a temperature-corrected threshold voltage value; Comparing the temperature-corrected threshold voltage value with a preset ideal threshold voltage model, calculating a deviation coefficient to obtain a correction coefficient, and calibrating the temperature-corrected threshold voltage value based on the correction coefficient to obtain a calibrated actual threshold voltage value; Calculating the difference between the actual threshold voltage value after calibration and the preset target threshold voltage value to obtain an initial threshold voltage deviation, and calculating the mean and standard deviation of the initial threshold voltage deviation to obtain a threshold voltage deviation distribution; Performing outlier detection based on the threshold voltage deviation distribution to obtain a valid threshold voltage deviation set, and performing weighted averaging on the valid threshold voltage deviation set to obtain a weighted average deviation value; The weighted average deviation value is compared with a preset allowable deviation range to determine whether it exceeds a threshold value, and a deviation degree determination result is obtained. Based on the deviation degree determination result, a threshold voltage deviation amount is calculated.

5. The MOS field effect tube threshold voltage adjustment method according to claim 1, characterized in that: The method of designing ion implantation parameters by using a particle swarm optimization algorithm according to the threshold voltage deviation, and regulating the structural parameters of the MOS field effect transistor to obtain the regulated structural parameters includes: The objective function of ion implantation is set based on the threshold voltage deviation, the deviation is converted into an optimization target of implantation dose and energy, a parameter optimization model is obtained, and the parameter optimization model is initialized to generate a particle group including implantation dose, implantation energy and implantation angle, and an initial particle group is obtained; Inputting the initial particle group into an ion implantation simulator to simulate the implantation process, obtaining an impurity distribution curve corresponding to each particle, and calculating the fitness value of each particle based on the impurity distribution curve, evaluating its effectiveness in adjusting the threshold voltage, and obtaining a particle fitness ranking; updating the optimal position of the particle swarm according to the fitness ranking of the particles, adjusting the speed and position of the particles to obtain an updated particle swarm, and repeatedly performing the injection simulation and fitness calculation on the updated particle swarm until a preset number of iterations is reached to obtain the optimal ion implantation parameters; Performing actual ion implantation on the MOS field effect transistor using the optimal ion implantation parameters, controlling the dose, energy and angle of the implantation equipment to obtain an implanted structure, and performing rapid thermal annealing on the implanted structure to activate the implanted impurity atoms and repair lattice defects to obtain a heat-treated structure; The key electrical parameters of the structure after the heat treatment, including threshold voltage, subthreshold swing and leakage current, are measured to obtain a structural parameter data set, and a device model is extracted based on the structural parameter data set to obtain the regulated structural parameters.

6. The MOS field effect tube threshold voltage adjustment method according to claim 1, characterized in that: The step of inputting the regulated structural parameters into a temperature sensitive circuit model, performing a multi-temperature threshold voltage simulation, and obtaining a temperature-threshold voltage response curve includes: Inputting the regulated structural parameters into a temperature-sensitive circuit model to obtain a temperature-dependent device model, and performing parameter extraction on the temperature-dependent device model to obtain a temperature-sensitive parameter set; Setting a plurality of temperature points based on the temperature sensitive parameter set to obtain a temperature point set, and performing circuit simulation on each temperature point in the temperature point set to obtain a voltage-current characteristic curve of each temperature point; Extracting the threshold voltage value of each temperature point from the voltage-current characteristic curve to obtain discrete temperature-threshold voltage data points, and performing data processing on the discrete temperature-threshold voltage data points to obtain a smoothed data set; The temperature-threshold voltage relationship is fitted based on the smoothed data set to obtain a response curve equation, and an error analysis is performed on the response curve equation to obtain an error range of the response curve, and a temperature-threshold voltage response curve is generated based on the response curve equation and the error range.

7. The MOS field effect tube threshold voltage adjustment method according to claim 1, characterized in that: The method of constructing an adaptive PID controller based on the temperature-threshold voltage response curve, dynamically compensating the threshold voltage of the MOS field effect tube, and outputting a real-time threshold voltage optimization strategy includes: Performing curve fitting on the temperature-threshold voltage response curve to obtain a functional relationship between the threshold voltage and the temperature, and calculating the temperature coefficient based on the functional relationship to obtain a sensitivity parameter of the threshold voltage changing with temperature; The sensitivity parameter is input into a PID controller, and the proportional, integral and differential parameters of the PID controller are preliminarily set to obtain an initial PID parameter group, and a feedback control loop is constructed based on the initial PID parameter group, and a threshold voltage measured in real time is compared with a target threshold voltage to obtain a threshold voltage deviation; Calculating the output signal of the PID controller according to the threshold voltage deviation to obtain a preliminary threshold voltage compensation value, and performing a limiting process on the preliminary threshold voltage compensation value to set a maximum compensation range to obtain an effective compensation signal; The effective compensation signal is converted into a gate voltage adjustment amount, and the gate voltage of the MOS field effect tube is adjusted to obtain a compensated threshold voltage, and the compensated threshold voltage is monitored in real time, and the compensation effect is recorded to obtain threshold voltage adjustment history data; Based on the threshold voltage adjustment historical data, the recursive least squares method is used to update the PID controller parameters to obtain optimized PID parameters, and the optimized PID parameters are applied to the controller, combined with the current temperature information, to output a real-time threshold voltage optimization strategy.

8. A MOS field effect tube threshold voltage adjustment device, characterized in that: The device is used to perform the MOS field effect tube threshold voltage adjustment method according to any one of claims 1 to 7, the device comprising: A scanning module is used to scan multiple groups of gate voltages and drain voltages of the MOS field effect transistor based on the initial structural parameters to obtain a multi-dimensional voltage-current characteristic data matrix; A fitting module, used for inputting the multi-dimensional voltage-current characteristic data matrix into a support vector regression model, performing feature extraction and fitting, and obtaining an initial threshold voltage function; A comparison module, used for calculating the actual threshold voltage value based on the initial threshold voltage function, and comparing it with the preset target threshold voltage value to obtain a threshold voltage deviation; A control module, used to design ion implantation parameters according to the threshold voltage deviation using a particle swarm optimization algorithm, and to control the structural parameters of the MOS field effect transistor to obtain the controlled structural parameters; A simulation module, used for inputting the regulated structural parameters into a temperature sensitive circuit model, performing multi-temperature point threshold voltage simulation, and obtaining a temperature-threshold voltage response curve; The output module is used to construct an adaptive PID controller based on the temperature-threshold voltage response curve, dynamically compensate the threshold voltage of the MOS field effect tube, and output a real-time threshold voltage optimization strategy.

9. A MOS field effect transistor threshold voltage adjustment device, characterized in that: The MOS field effect tube threshold voltage adjustment device comprises: a memory and at least one processor, wherein the memory stores instructions; The at least one processor calls the instruction in the memory so that the MOS field effect transistor threshold voltage adjustment device executes the MOS field effect transistor threshold voltage adjustment method according to any one of claims 1 to 7.

10. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instruction is executed by the processor, the MOS field effect transistor threshold voltage adjustment method according to any one of claims 1 to 7 is implemented.

Images