Multi-situation capacitance decision factor processing method based on big data and mathematical statistics

Abstract

The invention relates to the field of integrated circuit data processing, and discloses a multi-condition capacitance decision factor processing method based on big data and mathematical statistics. Comprising the following steps: collecting capacitance parameters and environment variable data, and generating a joint data set grouped according to environment parameter combination; based on environment parameter combination grouping, noise is eliminated, missing values are repaired, and an environment packetized multi-dimensional feature matrix is generated; analyzing principal components of the multi-dimensional feature matrix, and constructing a regression model to screen key factors; defining an environment parameter combination as a state space, and dynamically adjusting a weight coefficient; and loading the weight mapping table to a simulation platform to simulate capacitance characteristics under different environment parameter combinations, verifying the capacitance characteristics, and deploying the capacitance characteristics to a control module.

Description

Technical Field

[0001] The present invention relates to the technical field of integrated circuit software development data processing technology, and in particular to a multi-case capacitance decision factor processing method based on big data and mathematical statistics. Background Art

[0002] In the field of integrated circuit software development, the dynamic optimization of capacitor parameters directly affects the stability of circuit performance and environmental adaptability. In the existing technology, the extraction and optimization methods of capacitor decision factors under multiple environmental conditions usually rely on fixed thresholds or single environmental models, which makes it difficult to effectively quantify the complex nonlinear correlation between different environmental variables and capacitor characteristics. Especially when environmental parameters change dynamically, the existing methods lack the ability to mine the statistical characteristics of multi-dimensional capacitor data, resulting in rigid distribution of decision factor weights and insufficient model adaptability. For example, in scenarios with temperature fluctuations or unstable voltage, the existing methods cannot accurately identify the key influencing factors of capacitor parameters, resulting in reduced operating efficiency or abnormal response of integrated circuits. The above problems are caused by the insufficient ability of the data processing link to integrate multi-source heterogeneous capacitor features, and the lack of generalization ability of statistical models for dynamic environmental variables. It is urgent to build a unified cross-environment optimization framework through data-driven methods. Summary of the Invention

[0003] In response to the shortcomings of the existing technology, the present invention provides a multi-situation capacitance decision factor processing method based on big data and mathematical statistics. The present invention solves the problems of the capacitance decision factor extraction method based on fixed thresholds or single environmental models in the existing technology, which is difficult to quantify the capacitance characteristics and complex nonlinear correlation when multiple environmental variables change dynamically, resulting in rigid factor weight distribution and insufficient model environmental adaptability.

[0004] In order to solve the above technical problems, the specific technical solutions of the present invention are as follows: The present invention provides a multi-case capacitance decision factor processing method based on big data and mathematical statistics, including: Step 1: Collecting a capacitance parameter data set through a distributed sensor network, synchronously correlating it with environmental variable data, formatting the data in a unified manner, and generating a capacitance and environment joint data set grouped by environmental parameter combinations. Collecting a capacitance parameter data set in real time through a distributed sensor network, synchronously correlating it with dynamically changing environmental variable data, formatting the capacitance parameter data set and the environmental variable data in a unified manner, and generating a time-series correlated capacitance and environment joint data set. Step 2: Group the data set based on the combination of environmental parameters. The data set is grouped into groups according to temperature range, voltage level, and frequency range. Noise data is removed based on the outlier detection algorithm, and missing values are repaired through intra-group interpolation and filtering techniques to generate a multi-dimensional feature matrix for environmental grouping.

[0005] Step 3: Perform principal component analysis on the multidimensional feature matrix, calculate the eigenvectors through the covariance matrix, and extract the characteristic principal components whose cumulative variance contribution rate is higher than the preset threshold; construct a multivariate regression model based on the characteristic principal components, establish a nonlinear mapping relationship between capacitance parameters and environmental variables, screen key decision factors and generate an initial factor weight distribution table.

[0006] Step 4: Define the environmental parameter combination as a discrete state space, use the reinforcement learning algorithm to dynamically adjust the weight coefficients in the initial factor weight distribution table, and generate an environmental state and factor weight mapping table.

[0007] Step 5: Load the environmental state and factor weight mapping table into the integrated circuit simulation platform to simulate the capacitor response characteristics under different environmental parameter combinations. After verifying the model stability by comparing the test data, deploy the optimized weight parameters into the control module.

[0008] Furthermore, in the multi-case capacitance decision factor processing method based on big data and mathematical statistics of the present invention, step 1 includes: Collecting capacitance parameters and dynamically changing environmental variable data through a distributed sensor network, adding a unified timestamp to the capacitance parameters and the environmental variable data, and achieving time axis alignment across source data; Converting the analog signal in the time-axis aligned capacitance parameter data into a digital signal, and performing standardization processing on the digital signal to generate a structured data table including capacitance value, environmental parameter and timestamp fields; The structured data table is stored as a data grid of the capacitance and environment joint data set.

[0009] Furthermore, in the multi-case capacitance decision factor processing method based on big data and mathematical statistics of the present invention, step 2 includes: The generated joint data set of capacitance and environment is segmented into preset time windows, and the window size of the time windows is dynamically adjusted according to the frequency of changes in environmental parameters; Noise detection is performed on the segmented data set. Noise data is removed based on the outlier detection algorithm. Missing values are repaired through linear interpolation and smoothing filtering techniques of adjacent windows to generate a continuous time series data stream. The continuous time series data stream is multi-dimensionally aggregated according to the capacitance type and the environmental parameter category to form a multi-dimensional feature matrix with a row and column structure.

[0010] Furthermore, in the multi-case capacitance decision factor processing method based on big data and mathematical statistics of the present invention, step 3 includes: Calculate the covariance matrix of the multidimensional feature matrix to generate eigenvectors and eigenvalues; Extract the principal components with the cumulative variance contribution rate exceeding a preset threshold based on the feature vectors and eigenvalues; Input the principal components, capacitance parameter data, and environmental variable data into a multiple regression model to calculate the influence weights of each principal component on the capacitance performance; Screen the top N key decision factors in descending order according to the influence weights to generate an initial factor weight distribution table.

[0011] Further, in the multi - situation capacitance decision factor processing method based on big data and mathematical statistics of the present invention, step 4 includes: Divide the real - time collected environmental parameters into a state space according to a preset discretization interval, and associate the state space with the weight coefficients in the initial factor weight distribution table; Based on the Q - learning strategy, dynamically adjust the weight coefficients in the initial factor weight distribution table according to the state space corresponding to the current environmental parameters; Statistically calculate the deviation value between the real - time collected capacitance parameter data and the prediction result of the multiple regression model through a sliding window, and update the weight parameters in the environmental state - factor weight mapping table according to the deviation value.

[0012] Further, in the multi - situation capacitance decision factor processing method based on big data and mathematical statistics of the present invention, step 5 includes: Load the environmental state - factor weight mapping table into the integrated circuit simulation platform; Generate a random environmental fluctuation sequence based on the Monte Carlo method, simulate the extreme scenarios corresponding to the dynamically changing environmental variable data, and output the capacitance characteristic curve; Compare the error rate between the capacitance characteristic curve and the actual test data to evaluate the stability of the environmental state - factor weight mapping table; Deploy the dynamically adjusted weight parameters to the integrated circuit software control module through a lightweight API interface to complete the configuration of the dynamically adjusted weight parameters.

[0013] Further, the multi - situation capacitance decision factor processing method based on big data and mathematical statistics of the present invention further includes: The system adopts a version control mechanism to manage the environmental state - factor weight mapping table, retain historical weight configurations, and support incremental updates; After deploying the optimized weight parameters to the integrated circuit software control module, when a new capacitance parameter data set is collected subsequently, the system triggers processes including data cleaning, modeling, and dynamic adjustment, generates an updated environmental state - factor weight mapping table, and performs version control management on the updated mapping table to complete the continuous optimized deployment of the weight parameters.

[0014] Advantages of the present invention; The technical solution provided by the present invention constructs a multi-dimensional feature matrix through an environmental parameter combination grouping mechanism, combines principal component analysis to screen key decision factors, and breaks through the limitations of the fixed threshold model; uses a reinforcement learning algorithm to dynamically adjust the factor weight coefficients in the discrete state space to achieve real-time matching of environmental variable fluctuations and weight allocation; and based on the extreme scenario verification mechanism of Monte Carlo simulation, improves the generalization ability of capacitance characteristic prediction in a dynamic environment. Through version control and incremental update, a closed-loop optimization system is formed, enabling the weight mapping table to continuously adapt to changes in environmental parameters, effectively solving the problem of rigid factor weights caused by a single environmental model in the background technology, and improving the capacitance characteristic prediction accuracy and control stability of integrated circuits in multi-variable coupling scenarios of temperature, voltage, and frequency. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solution of the present invention, the drawings required for use in the embodiments will be briefly introduced below. Obviously, for those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0016] Figure 1 It is a flowchart of the multi-case capacitance decision factor processing method based on big data and mathematical statistics provided by the embodiments of the present invention.

[0017] Figure 2 It is a high-precision stability verification diagram provided by the embodiments of the present invention. [[ID=;15]]

[0018] Figure 3 It is a capacitance-voltage characteristic prediction accuracy diagram provided by the embodiments of the present invention.

[0019] Figure 4 It is a comparison diagram of model error rates under different environmental stresses provided by the embodiments of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0020] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with the specific embodiments and corresponding drawings of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of them. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention. The following will detail the technical solutions provided by each embodiment of the present invention in conjunction with the drawings. To better understand the objectives of the present invention, the present invention will be further described in detail below.

[0021] Please refer to Figures 1 to 4 , the multi-case capacitance decision factor processing method based on big data and mathematical statistics provided by the present invention includes: Step 1: Collecting a capacitance parameter data set through a distributed sensor network, synchronously correlating it with environmental variable data, formatting the data in a unified manner, and generating a capacitance and environment joint data set grouped by environmental parameter combinations. Collecting a capacitance parameter data set in real time through a distributed sensor network, synchronously correlating it with dynamically changing environmental variable data, formatting the capacitance parameter data set and the environmental variable data in a unified manner, and generating a time-series correlated capacitance and environment joint data set. Step 2: Group the data set based on the combination of environmental parameters. The data set is grouped into groups according to temperature range, voltage level, and frequency range. Noise data is removed based on the outlier detection algorithm, and missing values are repaired through intra-group interpolation and filtering techniques to generate a multi-dimensional feature matrix for environmental grouping.

[0022] Step 3: Perform principal component analysis on the multidimensional feature matrix, calculate the eigenvectors through the covariance matrix, and extract the characteristic principal components whose cumulative variance contribution rate is higher than the preset threshold; construct a multivariate regression model based on the characteristic principal components, establish a nonlinear mapping relationship between capacitance parameters and environmental variables, screen key decision factors and generate an initial factor weight distribution table.

[0023] Step 4: Define the environmental parameter combination as a discrete state space, use the reinforcement learning algorithm to dynamically adjust the weight coefficients in the initial factor weight distribution table, and generate an environmental state and factor weight mapping table.

[0024] Step 5: Load the environmental state and factor weight mapping table into the integrated circuit simulation platform to simulate the capacitor response characteristics under different environmental parameter combinations. After verifying the model stability by comparing the test data, deploy the optimized weight parameters into the control module.

[0025] Step 1: The distributed sensor network collects capacitor parameter datasets in real time, including physical quantities such as capacitance value and dielectric constant, and synchronously associates dynamically changing environmental variable data, including temperature, voltage, and frequency parameters. A unified timestamp is added to the multi-source data to solve the problem of sampling frequency differences across devices and achieve time axis alignment. The capacitor analog signal is converted into a digital signal through the analog-to-digital conversion module, and all digital signals are standardized to eliminate dimensional differences. The output is a structured data table containing timestamps, capacitance values, and environmental parameters, which are grouped by environmental parameter combinations and stored as a joint capacitor and environmental dataset, where the grouping is based on the combination of temperature range, voltage level, and frequency range.

[0026] Step 2: Based on the grouping results of environmental parameter combinations, adopt an event-triggered mechanism to dynamically segment the combined dataset: reset the time window when the change amount of environmental parameters exceeds a preset threshold. Identify and remove outlier noise data through the Isolation Forest algorithm, fill in missing values using Lagrangian interpolation of adjacent data windows, and then smooth the data through a Butterworth low-pass filter to generate a continuous time-series data stream. Perform multi-dimensional aggregation operations according to the capacitance type and environmental parameter categories to form an environmental grouped multi-dimensional feature matrix with time series as rows and environmental features as columns.

[0027] Step 3: Calculate the covariance matrix of the multi-dimensional feature matrix, and obtain the eigenvectors and eigenvalues through eigen-decomposition. Select the principal component vectors whose cumulative variance contribution rate exceeds the preset threshold, and the principal components represent the interaction effects between environmental variables. Input the principal component vectors, capacitance parameters, and environmental variables into a multiple regression model to establish a non-linear mapping relationship of capacitance values with respect to the principal components. Screen the top N key decision factors according to the absolute value ranking of the regression coefficients, and generate an initial factor weight distribution table containing factor names and weight coefficients.

[0028] Step 4: Discretize the environmental parameter combinations into a finite state space, and each state corresponds to a specific combination of temperature, voltage, and frequency intervals. Construct a state-action value function based on the Q-learning algorithm, and use the mean absolute error between the measured capacitance value and the predicted value of the regression model within the sliding window as the reward signal. Select the weight adjustment action through the greedy strategy, and dynamically update the coefficients in the initial factor weight distribution table to generate a dynamic mapping table of environmental states and factor weights.

[0029] Step 5: Load the environmental state and factor weight mapping table into the integrated circuit simulation platform, generate a random environmental fluctuation sequence based on the Monte Carlo method to simulate extreme scenarios. Output the capacitance characteristic simulation curves under different environmental parameter combinations, and calculate the mean relative error between the simulation results and the actual test data. When the error rate is lower than the preset tolerance threshold, write the optimized weight parameters into the configurable register of the control module through a lightweight API interface to complete parameter deployment.

[0030] The system adopts a version control mechanism to manage the historical versions of the environmental state-weight mapping table and supports incremental updates. After the weight parameter deployment, the newly collected capacitance parameter dataset triggers the data cleaning, modeling, and dynamic adjustment processes. Compare the differences between the output updated mapping table and the historical versions, and after regression test verification, deploy it to the control module through wireless update to form a continuous iterative optimization loop of weight parameters.

[0031] Specifically, for the multi-case capacitance decision factor processing method based on big data and mathematical statistics described in the present invention, the step 1 includes: Collect capacitance parameters and data of dynamically changing environmental variables through a distributed sensor network, add a unified timestamp to the capacitance parameters and the environmental variable data to achieve the time-axis alignment of cross-source data; Convert the analog signal in the capacitance parameter data after time-axis alignment into a digital signal, and perform standardization processing on the digital signal to generate a structured data table including capacitance values, environmental parameters, and timestamp fields; Store the structured data table as a data grid of the capacitance and environment joint dataset.

[0032] The distributed sensor network synchronously collects capacitance parameter data and environmental variable data. The capacitance parameter data includes physical quantities such as capacitance values and dielectric constants, and the environmental variable data includes temperature, voltage, and frequency parameters. To eliminate the sampling timing deviation of multi-source devices, a unified timestamp based on the NTP protocol is added to the two types of data to achieve time-axis alignment with millisecond-level accuracy. The capacitance analog signal after time-axis alignment is processed into a digital signal by the ADC conversion module, and the environmental variable data is directly subjected to digital standardization processing. The Z-score method is uniformly used to eliminate the dimension difference. The standardized data is reorganized into a structured data table according to the field type. The data table includes five core fields: timestamp, capacitance value, temperature value, voltage value, and frequency value. Finally, the structured data table is stored in the distributed database in the form of key-value pairs to form a data grid of the capacitance and environment joint dataset. The data grid establishes index partitions according to the three-dimensional combination of temperature range, voltage level, and frequency range.

[0033] Specifically, for the multi-case capacitance decision factor processing method based on big data and mathematical statistics described in the present invention, step 2 includes: Segment the generated capacitance and environment joint dataset according to a preset time window, and dynamically adjust the window size of the time window according to the change frequency of the environmental parameters; Perform noise detection on the segmented dataset, eliminate noise data based on the outlier detection algorithm, and repair missing values through linear interpolation and smoothing filtering techniques of adjacent windows to generate a continuous time-series data stream; Perform multi-dimensional aggregation on the continuous time-series data stream according to the capacitance type and environmental parameter category to form a multi-dimensional feature matrix with row and column structures.

[0034] The capacitance and environment combined dataset triggers a dynamic segmentation mechanism based on environmental parameter change events. When the change amount of temperature, voltage, or frequency exceeds a preset threshold, a new time window is generated. The window size is adaptively adjusted according to the fluctuation frequency of environmental parameters: in high-frequency fluctuation scenarios, the window is narrowed to capture transient features, and in low-frequency scenarios, the window is enlarged to enhance data continuity. The isolated forest algorithm is used to detect outliers in the segmented dataset, and noise data points beyond three standard deviation ranges are identified and removed. Missing values are repaired by using Lagrange interpolation of adjacent window data to fill in the missing positions, and then temporal smoothing processing is performed by a Butterworth low-pass filter to output a continuous and uninterrupted time series data stream. This data stream is classified according to the capacitance material type and aggregated and reorganized in three environmental parameter dimensions of temperature, voltage, and frequency to form a multi-dimensional feature matrix with timestamps as row indexes and combinations of capacitance types and environmental parameters as column features.

[0035] Specifically, for the method for processing multi-case capacitance decision factors based on big data and mathematical statistics described in the present invention, step 3 includes: Calculate the covariance matrix of the multi-dimensional feature matrix to generate eigenvectors and eigenvalues; Based on the eigenvectors and eigenvalues, extract the principal components whose cumulative variance contribution rate exceeds a preset threshold; Input the principal components, capacitance parameter data, and environmental variable data into a multiple regression model to calculate the influence weights of each principal component on the capacitance performance; According to the influence weights, screen the top N key decision factors in descending order to generate an initial factor weight distribution table.

[0036] Calculate the covariance matrix of the multi-dimensional feature matrix, and use the Jacobi iterative method to solve the eigenvectors and eigenvalues. Based on the sorting of the eigenvalue magnitudes, select the top K principal component vectors whose cumulative variance contribution rate exceeds a preset threshold. The principal components represent the interaction patterns among environmental variables. Input the principal component vectors, original capacitance parameter data, and environmental variable data into a multiple regression model, and use the least squares method to fit the mapping relationship between the principal components and the capacitance performance indicators. Calculate the standardized regression coefficients of each principal component in the regression model as their influence weight values on the capacitance performance. According to the influence weight values, sort them in descending order, and screen the environmental variables corresponding to the top N principal components with the highest weight values as the key decision factors. Output a set of key-value pairs containing factor names and weight coefficients to generate an initial factor weight distribution table.

[0037] Specifically, for the method for processing multi-case capacitance decision factors based on big data and mathematical statistics described in the present invention, step 4 includes: Divide the real-time collected environmental parameters into a state space according to a preset discretization interval, and associate the state space with the weight coefficients in the initial factor weight distribution table; Based on the Q-learning strategy, according to the state space corresponding to the current environmental parameters, dynamically adjust the weight coefficients in the initial factor weight distribution table; Statistically collect the deviation values between the real-time acquired capacitance parameter data and the prediction results of the multiple regression model through a sliding window, and update the weight parameters in the environmental state and factor weight mapping table according to the deviation values.

[0038] Divide the real-time acquired environmental parameters into a state space according to the preset discretization rules, where the temperature is discretized by Celsius intervals, the voltage is discretized by volt levels, and the frequency is discretized by Hertz frequency bands. Each discrete unit is defined as an independent state. Establish a mapping relationship between each state in the state space and the weight coefficients in the initial factor weight distribution table to form an initial state-weight association table. Based on the Q-learning algorithm, construct a state-action value function, where the action space is defined as operations of increasing, decreasing, or remaining unchanged of the weight coefficients. According to the discrete state corresponding to the current environmental parameters, adopt the ε-greedy strategy to select the weight adjustment action, and output the dynamically adjusted set of weight coefficients. Through the sliding statistical mechanism of a fixed time window, calculate the mean absolute deviation between the measured capacitance parameter values and the predicted values of the multiple regression model within the window. Use this deviation value as the immediate reward signal for Q-learning to update the weight parameter values in the environmental state and factor weight mapping table.

[0039] Specifically, for the multi-case capacitance decision factor processing method based on big data and mathematical statistics described in the present invention, step 5 includes: Load the environmental state and factor weight mapping table into the integrated circuit simulation platform; Generate a random environmental fluctuation sequence based on the Monte Carlo method, simulate the extreme scenarios corresponding to the dynamically changing environmental variable data, and output the capacitance characteristic curve; Compare the error rate between the capacitance characteristic curve and the actual test data, and evaluate the stability of the environmental state and factor weight mapping table; Deploy the dynamically adjusted weight parameters to the integrated circuit software control module through a lightweight API interface to complete the configuration of the dynamically adjusted weight parameters.

[0040] Load the environmental status and factor weight mapping table into the memory workspace of the SPICE integrated circuit simulation platform to establish a mapping relationship library between environmental parameters and capacitance response characteristics. Generate an environmental fluctuation sequence based on the Monte Carlo random sampling method, where temperature fluctuation is simulated using a random walk model, voltage fluctuation is simulated using a Poisson jump model, and frequency fluctuation is simulated using a Gaussian distribution model to generate a simulation input sequence containing extreme scenarios. The simulation platform executes circuit netlist calculations and outputs capacitance value-frequency characteristic curves under different combinations of environmental parameters. Calculate the average relative error between this characteristic curve and the actual test data set, and use whether the error rate is lower than a preset tolerance threshold as the evaluation criterion for the stability of the mapping table. Call the configuration register of the control module through a lightweight RESTful API interface, and write the dynamically adjusted weight parameters in JSON format to the non-volatile memory of the FPGA programmable logic unit to complete the configuration update of the dynamically adjusted weight parameters.

[0041] Specifically, the multi-case capacitance decision factor processing method based on big data and mathematical statistics described in the present invention further includes: The system uses a version control mechanism to manage the environmental status and factor weight mapping table, retains historical weight configurations, and supports incremental updates; After deploying the optimized weight parameters to the integrated circuit software control module, when a new capacitance parameter data set is collected subsequently, the system triggers processes including data cleaning, modeling, and dynamic adjustment, generates an updated environmental status and factor weight mapping table, and performs version control management on the updated mapping table to complete the continuous optimized deployment of the weight parameters.

[0042] The system uses a version control system based on branch management to manage the environmental status and factor weight mapping table. Each time the weight parameters are updated, an independent version snapshot is generated to retain a complete traceable record of the historical weight configurations. It supports an incremental update mechanism. By using a difference comparison algorithm to identify the weight coefficient change regions between the new and old mapping tables, only the changed parts are subjected to version storage operations. After the optimized weight parameters are deployed to the integrated circuit software control module, the system starts a data collection event listener. When a new capacitance parameter data set is collected by the subsequent distributed sensor network, it automatically triggers processes such as data cleaning, feature matrix construction, principal component analysis, and reinforcement learning dynamic adjustment. The updated mapping table output by the new process is merged with the currently effective version. After verifying the environmental adaptability through regression testing, the incremental update package is deployed to the firmware storage area of the control module via wireless OTA to complete the continuous iterative optimization of the weight parameters.

[0043] In the specific implementation of the present invention, a distributed sensor network is used to collect a capacitance parameter dataset and environmental variable data, where the environmental variable data includes temperature, voltage, and frequency parameters, and the capacitance parameter data includes capacitance values and dielectric constant physical quantities. A unified timestamp based on the NTP protocol is added to the multi-source data to achieve time-axis alignment with millisecond-level accuracy. The capacitance analog signal after time-axis alignment is processed into a digital signal by an ADC conversion module, and the environmental variable data is normalized using Z-score to eliminate the dimension difference, and a structured data table containing timestamps, capacitance values, and environmental parameters is output. This data table is indexed and partitioned according to the three-dimensional combination of temperature intervals, voltage levels, and frequency ranges, and stored as a data grid of a capacitance and environment joint dataset.

[0044] Based on environmental parameter change events, a dynamic segmentation mechanism is triggered. When the temperature change or voltage fluctuation exceeds a preset threshold, the time window is reset. The Isolation Forest algorithm is used to identify outlier noises beyond three standard deviations, and the missing values are filled by Lagrangian interpolation of adjacent windows, and a continuous time-series data stream is generated through a Butterworth low-pass filter. Classification is performed according to the capacitance material type and aggregation is performed in the dimension of environmental parameters to form a multi-dimensional feature matrix with rows corresponding to time series and columns corresponding to feature dimensions. The covariance matrix of the feature matrix is calculated, and the principal components with a cumulative variance contribution rate exceeding the threshold are extracted using the Jacobi iterative method. The standardized regression coefficients are calculated by inputting into a multiple regression model as influence weights, and the key decision factors are screened in descending order of the weight values to generate an initial weight distribution table.

[0045] The real-time environmental parameters are discretized into a state space according to Celsius intervals, volt levels, and hertz frequency bands, and each state is associated with the initial weight distribution table. A state-action value function is constructed based on the Q-learning algorithm, the increase and decrease operations of the weight coefficient are defined as the action space, and the ε-greedy strategy is used to select actions. The mean absolute deviation between the measured capacitance value and the model prediction value is statistically calculated through a fixed time window as the reward signal, and the environmental state and factor weight mapping table is dynamically updated. This mapping table is loaded into the SPICE simulation platform, and an extreme temperature fluctuation sequence is generated based on the Monte Carlo random walk model, and the capacitance value-frequency characteristic curve is output. The mean relative error between the curve and the actual test data is calculated. When the error rate is lower than the preset tolerance threshold, the weight parameters in JSON format are written into the non-volatile memory of the FPGA programmable logic unit through the RESTful API interface.

[0046] The system adopts a branch management mechanism to control the version of the mapping table, and generates an independent version snapshot each time an update is made. The newly collected capacitance parameter dataset triggers the entire process from data cleaning to dynamic adjustment, and after outputting the updated mapping table, a differential comparison algorithm is executed to identify the changed area. After verifying the environmental adaptability through regression testing, the incremental update package is deployed to the firmware storage area of the control module through OTA wireless update, forming a continuous closed-loop optimization mechanism for weight parameters, and solving the environmental adaptability defect of the static model in the background technology.

[0047] In the technical solution of the present invention, the environmental parameter combination grouping mechanism divides the data set through the three-dimensional combination of temperature range, voltage level, and frequency range, providing a structured input basis for subsequent noise elimination and feature extraction. The principal component analysis algorithm performs covariance calculation and eigenvalue decomposition on the multi-dimensional feature matrix, extracts the principal component vectors whose cumulative variance contribution rate exceeds the threshold, eliminates the interference of multicollinearity among environmental variables, and screens out the key influencing factors characterizing the capacitance performance. The multiple regression model establishes a non-linear mapping between the principal component vectors, the original capacitance parameters, and environmental variables, calculates the standardized regression coefficients through least squares fitting, and generates an initial factor weight distribution table.

[0048] The reinforcement learning algorithm adopts a discretized state space design, divides the real-time collected environmental parameters into discrete state units according to a preset interval, and each state is associated with an initial weight distribution table. The Q-learning strategy constructs a state-action value function, defines the increase, decrease, or maintenance of the weight coefficient as the action space, and selects actions to execute based on the ε-greedy strategy. The moving window statistics the mean absolute deviation between the measured capacitance value and the model predicted value as the immediate reward signal, driving the dynamic update of the weight parameters, and realizing the real-time matching of environmental state transition and weight allocation.

[0049] The Monte Carlo simulation method generates a temperature fluctuation sequence through a random walk model, simulates voltage fluctuations through a Poisson jump model, and constructs frequency perturbations through a Gaussian distribution model, covering extreme change scenarios of environmental parameters. The calculation of the error rate between the capacitance characteristic curve output by the simulation and the actual test data provides a quantitative basis for model stability. The version control mechanism adopts a branch management and differential comparison algorithm, only stores the incremental update package of the changed area of the mapping table, and forms a closed-loop optimization system in combination with regression testing and OTA wireless deployment. The cooperation of each algorithm realizes the transformation from static weight allocation to environment-adaptive decision-making, and solves the environmental generalization ability defect of the fixed threshold model.

[0050] The existing methods for extracting capacitance decision factors rely on fixed thresholds or single environmental models, and their defects stem from two aspects: First, the fixed threshold cannot adapt to the dynamic coupling effects of multiple environmental parameters such as temperature, voltage, and frequency, resulting in inaccurate quantification of the capacitance characteristics and non-linear correlation relationships; Second, the static weight allocation mechanism ignores the state space transition of the environmental parameter combinations. When the environmental variables fluctuate continuously, the factor weights cannot respond to the changes in real time, resulting in insufficient environmental adaptability of the model.

[0051] The present invention solves the above problems through the state space discretization of the environmental parameter combinations and the dynamic weight adjustment mechanism of reinforcement learning. First, the environmental parameter combinations are defined as a discrete state space, and an initial mapping relationship between the environmental state and the factor weights is established; Second, a state-action value function is constructed based on the Q-learning strategy, and the deviation between the measured capacitance value and the model prediction value is used as the reward signal to dynamically adjust the coefficients in the initial factor weight distribution table; Finally, an extreme environmental fluctuation sequence is generated through Monte Carlo simulation to verify the stability of the weight mapping table in a dynamic scenario. This mechanism enables the factor weight allocation to have environmental state self-adaptability.

[0052] The multi-dimensional feature matrix of environmental grouping provides input for principal component analysis, and key decision factors are extracted through covariance matrix calculation, breaking through the information limitation of the single environmental model; The dynamic weight adjustment process driven by reinforcement learning converts the changes in environmental parameters into state transition actions, realizing the real-time matching of weight coefficients and environmental fluctuations; The version control mechanism supports the incremental update of the mapping table, and combines the acquisition of new data to trigger a closed-loop optimization process, forming a continuously enhanced environmental generalization ability. Finally, dynamic weight parameters are deployed in the integrated circuit software control module to improve the prediction accuracy of capacitance characteristics.

Claims

1. A multi-case capacitance decision factor processing method based on big data and mathematical statistics, characterized in that include: Step 1: Collecting a capacitance parameter data set through a distributed sensor network, synchronously correlating it with environmental variable data, formatting the data in a unified manner, and generating a capacitance and environment joint data set grouped by environmental parameter combinations. Collecting a capacitance parameter data set in real time through a distributed sensor network, synchronously correlating it with dynamically changing environmental variable data, formatting the capacitance parameter data set and the environmental variable data in a unified manner, and generating a time-series correlated capacitance and environment joint data set. Step 2: Group the dataset based on the combination of environmental parameters. The dataset is grouped by temperature range, voltage level, and frequency range. Noise data is removed using an outlier detection algorithm, and missing values are repaired through intra-group interpolation and filtering techniques to generate a multi-dimensional feature matrix for environmental groupings. Step 3: Perform principal component analysis on the multidimensional feature matrix, calculate the eigenvectors through the covariance matrix, and extract the principal components with cumulative variance contribution rates higher than the preset threshold; A multivariate regression model is constructed based on the characteristic principal components, a nonlinear mapping relationship between capacitance parameters and environmental variables is established, key decision factors are screened, and an initial factor weight distribution table is generated; Step 4: Define the environmental parameter combination as a discrete state space, use the reinforcement learning algorithm to dynamically adjust the weight coefficients in the initial factor weight distribution table, and generate a mapping table between environmental state and factor weight; Step 5: Load the environmental state and factor weight mapping table into the integrated circuit simulation platform to simulate the capacitor response characteristics under different environmental parameter combinations. After verifying the model stability by comparing the test data, deploy the optimized weight parameters into the control module.

2. The multi-case capacitance decision factor processing method based on big data and mathematical statistics according to claim 1, characterized in that The step 1 comprises: Collecting capacitance parameters and dynamically changing environmental variable data through a distributed sensor network, adding a unified timestamp to the capacitance parameters and the environmental variable data, and achieving time axis alignment across source data; Converting the analog signal in the time-axis aligned capacitance parameter data into a digital signal, and performing standardization processing on the digital signal to generate a structured data table including capacitance value, environmental parameter and timestamp fields; The structured data table is stored as a data grid of the capacitance and environment joint data set.

3. The multi-case capacitance decision factor processing method based on big data and mathematical statistics according to claim 2, wherein The step 2 includes: The generated capacitance and environment joint data set is segmented according to a preset time window, and the window size of the time window is dynamically adjusted according to the frequency of change of the environmental parameters; Noise detection is performed on the segmented data set. Noise data is removed based on the outlier detection algorithm. Missing values are repaired through linear interpolation and smoothing filtering techniques of adjacent windows to generate a continuous time series data stream. The continuous time series data stream is multi-dimensionally aggregated according to the capacitance type and the environmental parameter category to form a multi-dimensional feature matrix with a row and column structure.

4. The multi-case capacitance decision factor processing method based on big data and mathematical statistics according to claim 3, characterized in that, The step 3 includes: Calculate the covariance matrix of the multidimensional feature matrix to generate eigenvectors and eigenvalues; Based on the eigenvectors and eigenvalues, extracting principal components whose cumulative variance contribution rate exceeds a preset threshold; Input the principal components, capacitance parameter data, and environmental variable data into a multiple regression model to calculate the influence weight of each principal component on capacitance performance; Screen the top N key decision factors according to the influence weights in descending order to generate an initial factor weight distribution table.

5. The method for processing multi-case capacitance decision factors based on big data and mathematical statistics according to claim 4, characterized in that Step 4 includes: Divide the real-time collected environmental parameters into a state space according to a preset discretization interval, and associate the state space with the weight coefficients in the initial factor weight distribution table; Based on the Q-learning strategy, dynamically adjust the weight coefficients in the initial factor weight distribution table according to the state space corresponding to the current environmental parameters; Statistically calculate the deviation value between the real-time collected capacitance parameter data and the prediction result of the multiple regression model through a sliding window, and update the weight parameters in the environmental state and factor weight mapping table according to the deviation value.

6. The method for processing multi - situation capacitance decision factors based on big data and mathematical statistics according to claim 5, wherein, Step 5 includes: Load the environmental state and factor weight mapping table into the integrated circuit simulation platform; Generate a random environmental fluctuation sequence based on the Monte Carlo method, simulate the extreme scenarios corresponding to the dynamically changing environmental variable data, and output the capacitance characteristic curve; Compare the error rate between the capacitance characteristic curve and the actual test data to evaluate the stability of the environmental state and factor weight mapping table; Deploy the dynamically adjusted weight parameters to the integrated circuit software control module through a lightweight API interface to complete the configuration of the dynamically adjusted weight parameters.

7. The multi-case capacitance decision factor processing method based on big data and mathematical statistics according to claim 6, wherein It also includes: The system uses a version control mechanism to manage the environmental state and factor weight mapping table, retains historical weight configurations and supports incremental updates; After deploying the optimized weight parameters to the integrated circuit software control module, when a new capacitance parameter data set is collected subsequently, the system triggers processes including data cleaning, modeling, and dynamic adjustment, generates an updated environmental state and factor weight mapping table, and performs version control management on the updated mapping table to complete the continuous optimized deployment of the weight parameters.

Images