Temperature adding instrument control system and method based on data analysis

Abstract

The invention discloses a temperature adding instrument control system and method based on data analysis, and relates to the technical field of data analysis. The system comprises a historical data collection and preprocessing module, a temperature-partial discharge characteristic correlation analysis module, a real-time demand analysis and matching module, a temperature addition strategy generation module and a real-time data analysis and feedback module. The historical data collection and preprocessing module collects historical data and constructs a historical feature data set; the temperature-partial discharge characteristic correlation analysis module establishes a correlation model of temperature change and partial discharge characteristics based on the historical characteristic data set; the real-time demand analysis and matching module performs analysis according to the real-time demand information of the partial discharge test to obtain a corresponding historical data unit; the temperature adding strategy generation module generates a preset temperature adding strategy according to the corresponding historical data in combination with the correlation model; and the real-time data analysis and feedback module evaluates the execution effect of the temperature adding strategy and adjusts the strategy through a feedback mechanism.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data analysis, in particular to a temperature increment instrument control system and method based on data analysis. BACKGROUND

[0002] In the power system, the insulation performance of power equipment is crucial. The partial discharge (PD) phenomenon is one of the key indicators reflecting the insulation condition of power equipment, and the four-in-one PD device is widely used in PD detection of power equipment. Since power equipment will face different temperature environments in actual operation, the change of temperature has a significant influence on the insulation performance and PD characteristics of power equipment. Therefore, during the test of the four-in-one PD device, accurately simulating and controlling the temperature conditions is of great significance for accurately evaluating the insulation state of power equipment.

[0003] However, as for the existing temperature control equipment, on the one hand, it is often difficult to meet the strict requirements of the four-in-one PD device test on temperature precision, control range, increment stability and compatibility with the PD device, etc., and there are many limitations, such as large temperature measurement error, inaccurate temperature control, inability to customize increment according to acceptance requirements, etc., which seriously affect the accuracy and reliability of the test results. On the other hand, the power acceptance requirements of each region often require increment according to the temperature value given by the acceptance department, and the existing temperature increment equipment cannot meet the requirement of arbitrary value increment. SUMMARY

[0004] The purpose of the present application is to provide a temperature increment instrument control system and method based on data analysis to solve the problems raised in the background.

[0005] In order to solve the above technical problems, the present application provides the following technical solutions: A temperature increment instrument control method based on data analysis, the method comprising the following steps: Step S100. Collecting temperature and PD signal historical data from PD test, the historical data including temperature data, environmental data, and PD signal records; pre-processing and feature extraction of the collected historical data, thereby obtaining feature parameters related to temperature change and PD characteristics, and constructing a historical feature data set; Step S200. According to the historical feature data set, aligning the feature parameters related to temperature change and PD characteristics in the historical data in time sequence, analyzing the correlation between temperature change and PD characteristics, and establishing a correlation model between temperature change and PD characteristics; Step S300. Input the real-time demand information of the partial discharge test, analyze the real-time demand information in combination with historical data to obtain historical data that matches the real-time demand information; and combine the correlation model between temperature change and partial discharge characteristics to generate a preset temperature increase strategy. Step S400. Collect real-time data of partial discharge test, and preprocess and extract features from the real-time data to obtain real-time characteristic parameters related to temperature change and partial discharge characteristics; evaluate the preset temperature increase strategy based on the real-time characteristic parameters, and perform corresponding processing according to the evaluation results.

[0006] Furthermore, step S100 includes: S101. Collect historical data of temperature and partial discharge signal from the partial discharge test. The historical data includes temperature data, environmental data, and partial discharge signal data. The temperature data is the temperature record at different time points during the partial discharge test. The environmental data is the environmental data of the partial discharge test, including temperature and humidity. The partial discharge signal data refers to data related to the partial discharge characteristics. The collected historical data is classified according to the corresponding partial discharge test cycle. Each partial discharge test cycle corresponds to a historical data unit, and each historical data unit includes historical data of one partial discharge test cycle. S102. Preprocess the historical data for each historical data unit. The preprocessing includes noise removal, data alignment, and normalization. The purpose of preprocessing is to remove noise, fill in missing values, unify the time scale of the data, and convert the data into a unified format. Feature extraction is performed on the preprocessed historical data. After completing the extraction of temperature and partial discharge signal features for each partial discharge test cycle, the feature parameters of temperature and partial discharge signal are arranged in chronological order to form a feature vector V containing multiple dimensions, where V = [v1, v2, ..., vs], where s represents the feature vector dimension, v1 represents the feature value corresponding to the first dimension of feature vector V, v2 represents the feature value corresponding to the second dimension of feature vector V, and so on, with vs representing the feature value corresponding to the s-th dimension of feature vector V. The feature vectors V of all historical data units are summarized to form a historical feature dataset.

[0007] Furthermore, the historical data for each historical data unit is preprocessed, and the specific details of data alignment during preprocessing are as follows: Assume the temperature data acquisition timestamp is t. a The corresponding temperature value is T a And a takes values ​​from 1 to n; assume the partial discharge signal data acquisition timestamp is t. b The corresponding partial discharge signal value is P b And b takes values ​​from 1 to m; where t b The interval ratio t aSmall, meaning m is much larger than n; for a given temperature data (t) a ,T a Construct a cubic spline interpolation function S(t) to represent every two adjacent timestamps t. a and t a+1 Temperature changes between: S a (t)=α a (tt a ) 3 +β a (tt a ) 2 +γ a (tt a )+φ; Among them, S a (t) represents the temperature interpolation function between the a-th and a+1-th t, where t∈[t a, t a+1 ],α a β a γ a And φ represents the spline interpolation coefficient; To ensure the smoothness of interpolation, the interpolation function must satisfy the following conditions: continuity of function values, continuity of the first derivative, and continuity of the second derivative, specifically expressed as: at adjacent data points t a and t a+1 At that point, the function values ​​are equal, i.e., S a (t a )=Ta and S a (t a+1 )=T a+1 ; at adjacent data points t a and t a+1 At that point, the derivative values ​​are equal, i.e., S' a (t a )=S' a-1 (t a ) and S' a (t a+1 )=S' a+1 (t a+1 ); at adjacent data point t a and t a+1 At this point, the second derivative values ​​are equal, i.e., S'' a (t a )=S'' a-1 (t a ) and S'' a (t a+1 )=S'' a+1 (t a+1 By constructing a spline polynomial for all adjacent data points, and satisfying the above continuity condition, the coefficient α can be obtained.a β a γ a And φ, the specific solution method usually uses a tridiagonal linear equation system to calculate these coefficients; for each partial discharge signal timestamp t b Find the temperature data timestamp interval [t] to which it belongs. a, t a+1 ], in the timestamp interval [t a, t a+1 Use the corresponding spline interpolation function S within ] a (t) is used to calculate the interpolated temperature value: T interp (t b )=S a (t b ).

[0008] Furthermore, step S200 includes: S201. Based on the historical feature dataset, align the feature parameters related to temperature changes and partial discharge characteristics in the historical data according to the time series to obtain {Tt(vi),Pt(vj)}, where Tt(vi) represents the i temperature change feature parameters corresponding to time point t, Pt(vj) represents the j feature parameters related to partial discharge characteristics corresponding to time point t, and i+j=s; calculate the Pearson correlation coefficient for each pair of feature parameters related to temperature changes and partial discharge characteristics, and compare the absolute value of the calculated Pearson correlation coefficient with the threshold ρ0; S202. If a Pearson correlation coefficient |ρ(T,P)| ≥ ρ0 exists, summarize the corresponding temperature changes and related characteristic parameters of partial discharge, and use a linear regression model to quantify the impact of temperature changes on the partial discharge signal. The linear regression model is as follows: Pt(vj)=w0+w1·Tt(v1)+w2·Tt(v2)+...+w i ·Tt(v i )+ϵ(t); Among them, Tt(v1),Tt(v2),...,Tt(v i ) represents the temperature change characteristic parameter linearly related to Pt(vj) corresponding to the partial discharge characteristics; ϵ(t) represents the error term of the linear regression model, w0, w1, w2, ..., w i Represents the regression coefficient; If a Pearson correlation coefficient |ρ(T,P)| < ρ0 exists, the corresponding characteristic parameters related to temperature changes and partial discharge characteristics are summarized and expressed using a support vector regression model, as follows: Pt(vj)=f[Tt(v1),Tt(v2),...,Tt(v i )]+G(t); Where f() represents the nonlinear mapping function, which is usually implemented using a kernel function; G(t) represents the error term of the support vector regression model; S203. Combining the linear regression model and the support vector regression model, a weighted fusion is performed to obtain a weighted fusion model, specifically expressed as follows: P fused (t)=λ·P linear (t)+(1-λ)·P SVR (t); Among them, P fused (t) represents the prediction result of the weighted fusion model, P linear (t) represents the prediction result of the linear regression model, P SVR (t) represents the prediction result of the support vector regression model, and λ represents the weight coefficient, which controls the balance between the two. It can usually be optimized by cross-validation.

[0009] Furthermore, step S300 includes: S301. Input the real-time requirements information for the partial discharge test, including the current state of the partial discharge test, the temperature control target, and the duration of the test; the current state of the partial discharge test includes the current partial discharge signal data and the current environmental data; compare each partial discharge test cycle in the historical feature dataset with the current real-time partial discharge signal, calculate the similarity, and based on the similarity calculation results, summarize the historical data units with similarity scores greater than or equal to the similarity threshold as the data units to be matched; obtain the environmental data of the data units to be matched, calculate the Euclidean distance between the environmental data of the data units to be matched and the current environmental data, and select the data units to be matched with the smallest Euclidean distance as the final matched historical data; S302. The predicted partial discharge signal P is obtained through a correlation model between temperature change and partial discharge characteristics. fused (t); Based on real-time demand information, obtain the current temperature T. current Target temperature T target Experiment duration T duration And the control target P of the partial discharge signal target (t); Calculate the difference between the target temperature and the current temperature: ΔT = T target -T current According to the duration T of the experiment duration Calculate the ideal rate of temperature change during the experiment: B_ideal = ΔT / T duration According to the predicted fusion result P fused (t), calculate the adjustment factor h(t), the specific calculation formula is: h(t) = 1 / [1 + u·|P fused (t)-P target(t)|];where u represents the parameter controlling the sensitivity adjustment;according to the adjustment factor h(t), the temperature increase rate B(t)=B_ideal·h(t) is calculated;according to the temperature increase rate B(t), the preset temperature increase strategy is obtained.

[0010] Furthermore, step S400 includes: S401. Conduct a partial discharge test according to the preset temperature increase strategy, collect real-time data of the partial discharge test, and preprocess and extract features from the real-time data according to the analysis method of historical data to obtain real-time characteristic parameters related to temperature change and partial discharge characteristics; calculate the mean square error R between the real-time characteristic parameters and the historical characteristic parameters corresponding to the historical data finally matched in S301, and the specific calculation formula is as follows: R=(1 / s)∑x∈[1,s](Fx_actual-Fx_history) 2 ; Where x represents the feature parameter number, Fx_actual represents the real-time feature parameter, and Fx_history represents the historical feature parameter corresponding to the final matched historical data; S402. Compare the calculated mean square error R with the threshold interval [R0, R1] to evaluate the preset temperature increase strategy; if the mean square error R belongs to the threshold interval [R0, R1], then retain the real-time data of the current partial discharge test; if the mean square error R does not belong to the threshold interval [R0, R1], then the relevant personnel shall make corresponding adjustments to the preset temperature increase strategy, and conduct the partial discharge test again according to the adjusted preset temperature increase strategy.

[0011] A temperature feeder control system based on data analysis includes: a historical data collection and preprocessing module, a temperature-partial discharge characteristic correlation analysis module, a real-time demand analysis and matching module, a temperature feed strategy generation module, and a real-time data analysis and feedback module; The historical data collection and preprocessing module collects historical data on temperature and partial discharge signals from the partial discharge test. The historical data includes temperature data, environmental data, and records of partial discharge signals. The collected historical data is preprocessed and features are extracted to obtain feature parameters related to temperature changes and partial discharge characteristics, and a historical feature dataset is constructed. The temperature-partial discharge characteristic correlation analysis module aligns the temperature changes and related characteristic parameters of the historical data according to the time series based on the historical feature dataset, analyzes the correlation between temperature changes and partial discharge characteristics, and thus establishes a correlation model between temperature changes and partial discharge characteristics. The real-time demand analysis and matching module takes into account the real-time demand information of the partial discharge test, analyzes the real-time demand information in combination with historical data, and obtains historical data that matches the real-time demand information. The temperature boosting strategy generation module combines the correlation model between temperature change and partial discharge characteristics to generate a preset temperature boosting strategy; The real-time data analysis and feedback module collects real-time data from the partial discharge test, and performs preprocessing and feature extraction on the real-time data to obtain real-time characteristic parameters related to temperature changes and partial discharge characteristics; it evaluates the preset temperature increase strategy based on the real-time characteristic parameters and performs corresponding processing based on the evaluation results.

[0012] Furthermore, the historical data collection and preprocessing module includes a historical data acquisition unit, a data preprocessing unit, and a feature extraction unit; The historical data acquisition unit is responsible for collecting historical data from the partial discharge test, including temperature data, environmental data, and partial discharge signal data; the data preprocessing unit preprocesses the collected historical data; the feature extraction unit extracts features from the preprocessed data to obtain key features of temperature and partial discharge signal, forms feature vectors, and constructs a historical feature dataset. The temperature-partial discharge characteristic correlation analysis module includes a time series alignment unit, a correlation analysis and modeling unit, and a model fusion and optimization unit. The time series alignment unit performs time series alignment on the characteristic parameters related to temperature changes and partial discharge characteristics based on historical feature datasets; the correlation analysis and modeling unit analyzes the relationship between temperature changes and partial discharge signal characteristics, and constructs linear regression and support vector regression models based on the correlation analysis results; the model fusion and optimization unit weighted and fused the prediction results of the linear regression model and the support vector regression model to form a weighted fusion model.

[0013] Furthermore, the real-time demand analysis and matching module includes a real-time demand information analysis unit and a historical data matching unit; The real-time demand information analysis unit takes into account the real-time demand information of the current partial discharge test and analyzes the similarity between the real-time demand information and historical data; the historical data matching unit obtains the final matched historical data based on the analysis results of the real-time demand information analysis unit.

[0014] Furthermore, the real-time data analysis and feedback module includes a real-time data collection and analysis unit and a real-time feedback unit; The real-time data collection and analysis unit collects real-time data from the partial discharge test, performs preprocessing and feature extraction on the real-time data in the same manner as historical data, and obtains real-time feature parameters. The real-time feedback unit analyzes the real-time feature parameters with the historical feature parameters corresponding to the finally matched historical data, calculates the mean square error, and compares it with a preset threshold range. If the mean square error is within the preset threshold range, the corresponding real-time data is saved. If the mean square error exceeds the preset threshold range, the corresponding feedback mechanism is triggered to adjust the temperature increase strategy.

[0015] Compared with the prior art, the beneficial effects of the present invention are: By establishing a historical feature dataset and combining it with a correlation model between temperature changes and partial discharge signal characteristics, this invention can accurately predict the impact of temperature changes on partial discharge signals, thereby achieving more precise temperature control during partial discharge testing. Compared to the problems of large temperature measurement errors and inaccurate temperature control in traditional temperature control equipment, this invention can automatically adjust the rate of temperature change and the escalation strategy through a data-driven approach, reducing temperature control fluctuations and ensuring that the dynamic changes of temperature and partial discharge signals remain consistent during the test.

[0016] This invention can generate a preset temperature increase strategy based on the real-time requirements of partial discharge testing and the specific temperature value requirements of the acceptance department, and automatically adjust the increase rate based on real-time data. Compared with the limitations of existing temperature increase equipment that cannot meet the requirement of arbitrary temperature increase, this invention can flexibly and precisely control the temperature increase according to different experimental needs, and meet the power acceptance requirements of different regions.

[0017] This invention establishes a correlation model between temperature changes and partial discharge signals through in-depth analysis of historical data, and uses data analysis methods such as Pearson correlation coefficient, linear regression model, and support vector regression model to quantify the impact of temperature on partial discharge signals. This data analysis-based approach can effectively identify the nonlinear relationship between temperature changes and partial discharge signals, thereby achieving more precise control of the partial discharge test process.

[0018] By collecting partial discharge test data in real time and comparing and matching it with historical data, the mean square error between real-time characteristic parameters and historical data is calculated. This invention can dynamically evaluate the effectiveness of the temperature boosting strategy. If the error exceeds a predetermined range, the system can automatically alert and adjust the temperature boosting strategy to ensure experimental accuracy and reliability. This feedback mechanism can improve the accuracy of experimental results and avoid adverse consequences caused by control errors. Attached Figure Description

[0019] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a schematic diagram of a temperature feeder control system module based on data analysis according to the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Please see Figure 1 The present invention provides the following technical solution: A temperature feeder control system based on data analysis includes: a historical data collection and preprocessing module, a temperature-partial discharge characteristic correlation analysis module, a real-time demand analysis and matching module, a temperature feed strategy generation module, and a real-time data analysis and feedback module; The historical data collection and preprocessing module collects historical data on temperature and partial discharge signals from the partial discharge test. The historical data includes temperature data, environmental data, and records of partial discharge signals. The collected historical data is preprocessed and features are extracted to obtain feature parameters related to temperature changes and partial discharge characteristics, and a historical feature dataset is constructed. The temperature-partial discharge characteristic correlation analysis module aligns the temperature changes and related characteristic parameters of the historical data according to the time series based on the historical feature dataset, analyzes the correlation between temperature changes and partial discharge characteristics, and thus establishes a correlation model between temperature changes and partial discharge characteristics. The real-time demand analysis and matching module takes into account the real-time demand information of the partial discharge test, analyzes the real-time demand information in combination with historical data, and obtains historical data that matches the real-time demand information. The temperature boosting strategy generation module combines the correlation model between temperature change and partial discharge characteristics to generate a preset temperature boosting strategy; The real-time data analysis and feedback module collects real-time data from the partial discharge test, and performs preprocessing and feature extraction on the real-time data to obtain real-time characteristic parameters related to temperature changes and partial discharge characteristics; it evaluates the preset temperature increase strategy based on the real-time characteristic parameters and performs corresponding processing based on the evaluation results.

[0022] The historical data collection and preprocessing module includes a historical data acquisition unit, a data preprocessing unit, and a feature extraction unit; The historical data acquisition unit is responsible for collecting historical data from the partial discharge test, including temperature data, environmental data, and partial discharge signal data; the data preprocessing unit preprocesses the collected historical data; the feature extraction unit extracts features from the preprocessed data to obtain key features of temperature and partial discharge signal, forms feature vectors, and constructs a historical feature dataset. The temperature-partial discharge characteristic correlation analysis module includes a time series alignment unit, a correlation analysis and modeling unit, and a model fusion and optimization unit. The time series alignment unit performs time series alignment on the characteristic parameters related to temperature changes and partial discharge characteristics based on historical feature datasets; the correlation analysis and modeling unit analyzes the relationship between temperature changes and partial discharge signal characteristics, and constructs linear regression and support vector regression models based on the correlation analysis results; the model fusion and optimization unit weighted and fused the prediction results of the linear regression model and the support vector regression model to form a weighted fusion model.

[0023] The real-time demand analysis and matching module includes a real-time demand information analysis unit and a historical data matching unit; The real-time demand information analysis unit takes into account the real-time demand information of the current partial discharge test and analyzes the similarity between the real-time demand information and historical data; the historical data matching unit obtains the final matched historical data based on the analysis results of the real-time demand information analysis unit.

[0024] The real-time data analysis and feedback module includes a real-time data collection and analysis unit and a real-time feedback unit; The real-time data collection and analysis unit collects real-time data from the partial discharge test, performs preprocessing and feature extraction on the real-time data in the same manner as historical data, and obtains real-time feature parameters. The real-time feedback unit analyzes the real-time feature parameters with the historical feature parameters corresponding to the finally matched historical data, calculates the mean square error, and compares it with a preset threshold range. If the mean square error is within the preset threshold range, the corresponding real-time data is saved. If the mean square error exceeds the preset threshold range, the corresponding feedback mechanism is triggered to adjust the temperature increase strategy.

[0025] A data analysis-based control method for a temperature feeder includes the following steps: Step S100. Collect historical data on temperature and partial discharge signal from the partial discharge test. The historical data includes temperature data, environmental data, and records of partial discharge signal. Preprocess and extract features from the collected historical data to obtain feature parameters related to temperature changes and partial discharge characteristics, and construct a historical feature dataset. Step S200. Based on the historical feature dataset, align the feature parameters related to temperature changes and partial discharge characteristics in the historical data according to the time series, analyze the correlation between temperature changes and partial discharge characteristics, and thus establish a correlation model between temperature changes and partial discharge characteristics. Step S300. Input the real-time demand information of the partial discharge test, analyze the real-time demand information in combination with historical data to obtain historical data that matches the real-time demand information; and combine the correlation model between temperature change and partial discharge characteristics to generate a preset temperature increase strategy. Step S400. Collect real-time data of partial discharge test, and preprocess and extract features from the real-time data to obtain real-time characteristic parameters related to temperature change and partial discharge characteristics; evaluate the preset temperature increase strategy based on the real-time characteristic parameters, and perform corresponding processing according to the evaluation results.

[0026] Step S100 includes: S101. Collect historical data of temperature and partial discharge signal from the partial discharge test. The historical data includes temperature data, environmental data, and partial discharge signal data. The temperature data is the temperature record at different time points during the partial discharge test. The environmental data is the environmental data of the partial discharge test, including temperature and humidity. The partial discharge signal data refers to data related to the partial discharge characteristics. The collected historical data is classified according to the corresponding partial discharge test cycle. Each partial discharge test cycle corresponds to a historical data unit, and each historical data unit includes historical data of one partial discharge test cycle. S102. Preprocess the historical data for each historical data unit. The preprocessing includes noise removal, data alignment, and normalization. The purpose of preprocessing is to remove noise, fill in missing values, unify the time scale of the data, and convert the data into a unified format. Feature extraction is performed on the preprocessed historical data. After completing the extraction of temperature and partial discharge signal features for each partial discharge test cycle, the feature parameters of temperature and partial discharge signal are arranged in chronological order to form a feature vector V containing multiple dimensions, where V = [v1, v2, ..., vs], where s represents the feature vector dimension, v1 represents the feature value corresponding to the first dimension of feature vector V, v2 represents the feature value corresponding to the second dimension of feature vector V, and so on, with vs representing the feature value corresponding to the s-th dimension of feature vector V. The feature vectors V of all historical data units are summarized to form a historical feature dataset.

[0027] In this embodiment, the feature extraction process for historical data is as follows: For the characteristic parameters corresponding to the temperature data, a corresponding temperature change trend curve L1 is plotted based on the temperature data during the partial discharge test cycle. The horizontal axis of the temperature change trend curve L1 represents the time points during the partial discharge test cycle, and the vertical axis represents the temperature data. Based on the temperature change trend curve L1: assuming the average temperature T is extracted... mean Temperature fluctuation range T range and the rate of temperature change T rate Wherein, the temperature fluctuation amplitude T range The temperature change rate T is the difference between the maximum and minimum temperatures during the partial discharge test cycle. rate This represents the rate of temperature change over time, and the formula for calculation is: T rate (t)=[T(t)-T(t-1)] / Δt, where T(t) is the temperature at time t, T(t-1) is the temperature at time t-1, and Δt is the time interval; Based on the characteristic parameters corresponding to the partial discharge signal data, time-domain and frequency-domain analyses are performed on the partial discharge signal data. It is assumed that the time-domain characteristics include: the peak value P of the partial discharge signal. peak Mean value of partial discharge signal P mean The variance of the partial discharge signal, P(σ) 2 ); Assume the frequency domain characteristics include: dominant frequency Pf peak It is usually extracted through Fourier transform; spectral energy PE total , represents the energy distribution of the partial discharge signal in the frequency domain, and is usually calculated by measuring the total spectral energy of the signal or the energy distribution of each frequency band; frequency bandwidth PBW represents the frequency distribution range of the partial discharge signal, and is usually determined by calculating the half-power point of the spectrum.

[0028] For each historical data unit, the historical data is preprocessed, and the specific details of data alignment during preprocessing are as follows: Assume the temperature data acquisition timestamp is t. a The corresponding temperature value is T a And a takes values ​​from 1 to n; assume the partial discharge signal data acquisition timestamp is t. b The corresponding partial discharge signal value is P b And b takes values ​​from 1 to m; where t b The interval ratio t a Small, meaning m is much larger than n; for a given temperature data (t) a ,T a Construct a cubic spline interpolation function S(t) to represent every two adjacent timestamps t. a and t a+1 Temperature changes between: S a (t)=α a (tt a) 3 +β a (tt a ) 2 +γ a (tt a )+φ; Among them, S a (t) represents the temperature interpolation function between the a-th and a+1-th t, where t∈[t a, t a+1 ],α a β a γ a And φ represents the spline interpolation coefficient; To ensure the smoothness of interpolation, the interpolation function must satisfy the following conditions: continuity of function values, continuity of the first derivative, and continuity of the second derivative, specifically expressed as: at adjacent data points t a and t a+1 At that point, the function values ​​are equal, i.e., S a (t a )=Ta and S a (t a+1 )=T a+1 ; at adjacent data points t a and t a+1 At that point, the derivative values ​​are equal, i.e., S' a (t a )=S' a-1 (t a ) and S' a (t a+1 )=S' a+1 (t a+1 ); at adjacent data point t a and t a+1 At this point, the second derivative values ​​are equal, i.e., S'' a (t a )=S'' a-1 (t a ) and S'' a (t a+1 )=S'' a+1 (t a+1 By constructing a spline polynomial for all adjacent data points, and satisfying the above continuity condition, the coefficient α can be obtained. a β a γ a And φ, the specific solution method usually uses a tridiagonal linear equation system to calculate these coefficients; for each partial discharge signal timestamp t b Find the temperature data timestamp interval [t] to which it belongs. a, t a+1 ], in the timestamp interval [t a, t a+1Use the corresponding spline interpolation function S within ] a (t) is used to calculate the interpolated temperature value: T interp (t b )=S a (t b ).

[0029] Step S200 includes: S201. Based on the historical feature dataset, align the feature parameters related to temperature changes and partial discharge characteristics in the historical data according to the time series to obtain {Tt(vi),Pt(vj)}, where Tt(vi) represents the i temperature change feature parameters corresponding to time point t, Pt(vj) represents the j feature parameters related to partial discharge characteristics corresponding to time point t, and i+j=s; calculate the Pearson correlation coefficient for each pair of feature parameters related to temperature changes and partial discharge characteristics, and compare the absolute value of the calculated Pearson correlation coefficient with the threshold ρ0; S202. If a Pearson correlation coefficient |ρ(T,P)| ≥ ρ0 exists, summarize the corresponding temperature changes and related characteristic parameters of partial discharge, and use a linear regression model to quantify the impact of temperature changes on the partial discharge signal. The linear regression model is as follows: Pt(vj)=w0+w1·Tt(v1)+w2·Tt(v2)+...+w i ·Tt(v i )+ϵ(t); Among them, Tt(v1),Tt(v2),...,Tt(v i ) represents the temperature change characteristic parameter linearly related to Pt(vj) corresponding to the partial discharge characteristics; ϵ(t) represents the error term of the linear regression model, w0, w1, w2, ..., w i Represents the regression coefficient; If a Pearson correlation coefficient |ρ(T,P)| < ρ0 exists, the corresponding characteristic parameters related to temperature changes and partial discharge characteristics are summarized and expressed using a support vector regression model, as follows: Pt(vj)=f[Tt(v1),Tt(v2),...,Tt(v i )]+G(t); Where f() represents the nonlinear mapping function, which is usually implemented using a kernel function; G(t) represents the error term of the support vector regression model; S203. Combining the linear regression model and the support vector regression model, a weighted fusion is performed to obtain a weighted fusion model, specifically expressed as follows: P fused (t)=λ·P linear (t)+(1-λ)·P SVR (t); Among them, P fused (t) represents the prediction result of the weighted fusion model, P linear (t) represents the prediction result of the linear regression model, P SVR (t) represents the prediction result of the support vector regression model, and λ represents the weight coefficient, which controls the balance between the two. It can usually be optimized by cross-validation.

[0030] Step S300 includes: S301. Input the real-time requirements information for the partial discharge test, including the current state of the partial discharge test, the temperature control target, and the duration of the test; the current state of the partial discharge test includes the current partial discharge signal data and the current environmental data; compare each partial discharge test cycle in the historical feature dataset with the current real-time partial discharge signal, calculate the similarity, and based on the similarity calculation results, summarize the historical data units with similarity scores greater than or equal to the similarity threshold as the data units to be matched; obtain the environmental data of the data units to be matched, calculate the Euclidean distance between the environmental data of the data units to be matched and the current environmental data, and select the data units to be matched with the smallest Euclidean distance as the final matched historical data; S302. The predicted partial discharge signal P is obtained through a correlation model between temperature change and partial discharge characteristics. fused (t); Based on real-time demand information, obtain the current temperature T. current Target temperature T target Experiment duration T duration And the control target P of the partial discharge signal target (t); Calculate the difference between the target temperature and the current temperature: ΔT = T target -T current According to the duration T of the experiment duration Calculate the ideal rate of temperature change during the experiment: B_ideal = ΔT / T duration According to the predicted fusion result P fused (t), calculate the adjustment factor h(t), the specific calculation formula is: h(t) = 1 / [1 + u·|P fused (t)-P target (t)|];where u represents the parameter controlling the sensitivity adjustment;according to the adjustment factor h(t), the temperature increase rate B(t)=B_ideal·h(t) is calculated;according to the temperature increase rate B(t), the preset temperature increase strategy is obtained.

[0031] In this embodiment, it is assumed that the difference between the target temperature and the current temperature is: ΔT = T target -T current =50-20=30℃; Assume the experiment duration is T. duration=120 minutes, calculate the ideal rate of temperature change B_ideal: B_ideal=ΔT / T duration =30 / 120=0.25℃ / min; Assume the correlation model derives the predicted partial discharge signal P based on historical data and temperature variations. fused Given that (t) = 0.5, calculate the corresponding adjustment factor h(t), where h(t) = 1 / [1 + u·|P fused (t)-P target [(t)|], assuming the control adjustment sensitivity u=0.1, the target partial discharge signal P target If (t) = 0.55, then h(t) = 0.995; calculate the actual temperature increment rate according to the adjustment factor: B(t) = B_ideal·h(t) = 0.25 × 0.995 = 0.24875 ℃ / min; thus, the preset temperature increment strategy is: in the next 120 minutes, the system should gradually increase the temperature according to this temperature increment rate so that the final temperature reaches the target temperature of 50℃.

[0032] Step S400 includes: S401. Conduct a partial discharge test according to the preset temperature increase strategy, collect real-time data of the partial discharge test, and preprocess and extract features from the real-time data according to the analysis method of historical data to obtain real-time characteristic parameters related to temperature change and partial discharge characteristics; calculate the mean square error R between the real-time characteristic parameters and the historical characteristic parameters corresponding to the historical data finally matched in S301, and the specific calculation formula is as follows: R=(1 / s)∑x∈[1,s](Fx_actual-Fx_history) 2 ; Where x represents the feature parameter number, Fx_actual represents the real-time feature parameter, and Fx_history represents the historical feature parameter corresponding to the final matched historical data; S402. Compare the calculated mean square error R with the threshold interval [R0, R1] to evaluate the preset temperature increase strategy; if the mean square error R belongs to the threshold interval [R0, R1], then retain the real-time data of the current partial discharge test; if the mean square error R does not belong to the threshold interval [R0, R1], then the relevant personnel shall make corresponding adjustments to the preset temperature increase strategy, and conduct the partial discharge test again according to the adjusted preset temperature increase strategy.

[0033] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0034] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A temperature feeder control method based on data analysis, characterized in that: The method includes the following steps: Step S100. Collect historical data on temperature and partial discharge signal from the partial discharge test. The historical data includes temperature data, environmental data, and records of partial discharge signal. Preprocess and extract features from the collected historical data to obtain feature parameters related to temperature changes and partial discharge characteristics, and construct a historical feature dataset. Step S200. Based on the historical feature dataset, align the feature parameters related to temperature changes and partial discharge characteristics in the historical data according to the time series, analyze the correlation between temperature changes and partial discharge characteristics, and thus establish a correlation model between temperature changes and partial discharge characteristics. Step S300. Input the real-time demand information of the partial discharge test, analyze the real-time demand information in combination with historical data to obtain historical data that matches the real-time demand information; and combine the correlation model between temperature change and partial discharge characteristics to generate a preset temperature increase strategy. Step S400. Collect real-time data of partial discharge test, and preprocess and extract features from the real-time data to obtain real-time characteristic parameters related to temperature change and partial discharge characteristics; evaluate the preset temperature increase strategy based on the real-time characteristic parameters, and perform corresponding processing according to the evaluation results.

2. The temperature feeder control method based on data analysis according to claim 1, characterized in that: Step S100 includes: S101. Collect historical data of temperature and partial discharge signal from the partial discharge test. The historical data includes temperature data, environmental data, and partial discharge signal data. The temperature data is the temperature record at different time points during the partial discharge test. The environmental data is the environmental data of the partial discharge test. The partial discharge signal data refers to data related to the partial discharge characteristics. The collected historical data is classified according to the corresponding partial discharge test cycle. Each partial discharge test cycle corresponds to a historical data unit, and each historical data unit includes historical data of one partial discharge test cycle. S102. Preprocess the historical data for each historical data unit, including noise removal, data alignment, and normalization; extract features from the preprocessed historical data; after extracting the temperature and partial discharge signal features for each partial discharge test cycle, arrange the feature parameters of the temperature and partial discharge signal in chronological order to form a feature vector V containing multiple dimensions, where V=[v1,v2,...,vs], where s represents the feature vector dimension, v1 represents the feature value corresponding to the first dimension of feature vector V, v2 represents the feature value corresponding to the second dimension of feature vector V, and so on, with vs representing the feature value corresponding to the s-th dimension of feature vector V; summarize the feature vectors V of all historical data units to form a historical feature dataset.

3. The temperature feeder control method based on data analysis according to claim 2, characterized in that: The specific details of the preprocessing of historical data for each historical data unit, including data alignment during preprocessing, are as follows: Assume the temperature data acquisition timestamp is t. a The corresponding temperature value is T a And a takes values ​​from 1 to n; assume the partial discharge signal data acquisition timestamp is t. b The corresponding partial discharge signal value is P b And b takes values ​​from 1 to m; where t b The interval ratio t a Small, meaning m is much larger than n; for a given temperature data (t) a ,T a Construct a cubic spline interpolation function S(t) to represent every two adjacent timestamps t. a and t a+1 Temperature changes between: S a (t)=α a (t-t a ) 3 +β a (t-t a ) 2 +γ a (t-t a )+φ; Among them, S a (t) represents the temperature interpolation function between the a-th and a+1-th t, where t∈[t a, t a+1 ],α a β a γ a And φ represents the spline interpolation coefficient; To ensure the smoothness of interpolation, the interpolation function must satisfy the following conditions: continuity of function values, continuity of the first derivative, and continuity of the second derivative, specifically expressed as: at adjacent data points t a and t a+1 At that point, the function values ​​are equal, i.e., S a (t a )=Ta and S a (t a+1 )=T a+1 ; at adjacent data points t a and t a+1 At that point, the derivative values ​​are equal, i.e., S' a (t a )=S' a-1 (t a ) and S' a (t a+1 )=S' a+1 (t a+1 ); at adjacent data point t a and t a+1 At this point, the second derivative values ​​are equal, i.e., S'' a (t a )=S'' a-1 (t a ) and S'' a (t a+1 )=S'' a+1 (t a+1 By constructing a spline polynomial for all adjacent data points, and satisfying the above continuity condition, the coefficient α can be obtained. a β a γ a And φ; for each partial discharge signal timestamp t b Find the temperature data timestamp interval [t] to which it belongs. a, t a+1 ], in the timestamp interval [t a, t a+1 Use the corresponding spline interpolation function S within ] a (t) is used to calculate the interpolated temperature value: T interp (t b )=S a (t b ).

4. The temperature feeder control method based on data analysis according to claim 1, characterized in that: Step S200 includes: S201. Based on the historical feature dataset, align the feature parameters related to temperature changes and partial discharge characteristics in the historical data according to the time series to obtain {Tt(vi),Pt(vj)}, where Tt(vi) represents the i temperature change feature parameters corresponding to time point t, Pt(vj) represents the j feature parameters related to partial discharge characteristics corresponding to time point t, and i+j=s; calculate the Pearson correlation coefficient for each pair of feature parameters related to temperature changes and partial discharge characteristics, and compare the absolute value of the calculated Pearson correlation coefficient with the threshold ρ0; S202. If a Pearson correlation coefficient |ρ(T,P)| ≥ ρ0 exists, summarize the corresponding temperature changes and related characteristic parameters of partial discharge, and use a linear regression model to quantify the impact of temperature changes on the partial discharge signal. The linear regression model is as follows: Pt(vj)=w0+w1·Tt(v1)+w2·Tt(v2)+...+w i ·Tt(v i )+ϵ(t); Among them, Tt(v1),Tt(v2),...,Tt(v i ) represents the temperature change characteristic parameter linearly related to Pt(vj) corresponding to the partial discharge characteristics; ϵ(t) represents the error term of the linear regression model, w0, w1, w2, ..., w i Represents the regression coefficient; If a Pearson correlation coefficient |ρ(T,P)| < ρ0 exists, the corresponding characteristic parameters related to temperature changes and partial discharge characteristics are summarized and expressed using a support vector regression model, as follows: Pt(vj)=f[Tt(v1),Tt(v2),...,Tt(v i )]+G(t); Where f() represents the nonlinear mapping function, which is usually implemented using a kernel function; G(t) represents the error term of the support vector regression model; S203. Combining the linear regression model and the support vector regression model, a weighted fusion is performed to obtain a weighted fusion model, specifically expressed as follows: P fused (t)=λ·P linear (t)+(1-λ)·P SVR (t); Among them, P fused (t) represents the prediction result of the weighted fusion model, P linear (t) represents the prediction result of the linear regression model, P SVR (t) represents the prediction result of the support vector regression model, and λ represents the weight coefficient.

5. The temperature feeder control method based on data analysis according to claim 4, characterized in that: Step S300 includes: S301. Input the real-time requirements information for the partial discharge test, including the current state of the partial discharge test, the temperature control target, and the duration of the test; the current state of the partial discharge test includes the current partial discharge signal data and the current environmental data; compare each partial discharge test cycle in the historical feature dataset with the current real-time partial discharge signal, calculate the similarity, and based on the similarity calculation results, summarize the historical data units with similarity scores greater than or equal to the similarity threshold as the data units to be matched; obtain the environmental data of the data units to be matched, calculate the Euclidean distance between the environmental data of the data units to be matched and the current environmental data, and select the data units to be matched with the smallest Euclidean distance as the final matched historical data; S302. The predicted partial discharge signal P is obtained through a correlation model between temperature change and partial discharge characteristics. fused (t); Based on real-time demand information, obtain the current temperature T. current Target temperature T target Experiment duration T duration And the control target P of the partial discharge signal target (t); Calculate the difference between the target temperature and the current temperature: ΔT = T target -T current According to the duration T of the experiment duration Calculate the ideal rate of temperature change during the experiment: B_ideal = ΔT / T duration According to the predicted fusion result P fused (t), calculate the adjustment factor h(t), the specific calculation formula is: h(t) = 1 / [1 + u·|P fused (t)-P target (t)|];where u represents the parameter controlling the sensitivity adjustment;according to the adjustment factor h(t), the temperature increase rate B(t)=B_ideal·h(t) is calculated;according to the temperature increase rate B(t), the preset temperature increase strategy is obtained.

6. The temperature feeder control method based on data analysis according to claim 5, characterized in that: Step S400 includes: S401. Conduct a partial discharge test according to the preset temperature increase strategy, collect real-time data of the partial discharge test, and preprocess and extract features from the real-time data according to the analysis method of historical data to obtain real-time characteristic parameters related to temperature change and partial discharge characteristics; calculate the mean square error R between the real-time characteristic parameters and the historical characteristic parameters corresponding to the historical data finally matched in S301, and the specific calculation formula is as follows: R=(1 / s)∑x∈[1,s](Fx_actual-Fx_history) 2 ; Where x represents the feature parameter number, Fx_actual represents the real-time feature parameter, and Fx_history represents the historical feature parameter corresponding to the final matched historical data; S402. Compare the calculated mean square error R with the threshold interval [R0, R1] to evaluate the preset temperature increase strategy; if the mean square error R belongs to the threshold interval [R0, R1], then retain the real-time data of the current partial discharge test; if the mean square error R does not belong to the threshold interval [R0, R1], then the relevant personnel shall make corresponding adjustments to the preset temperature increase strategy, and conduct the partial discharge test again according to the adjusted preset temperature increase strategy.

7. A temperature feeder control system based on data analysis, applied to the temperature feeder control method based on data analysis according to any one of claims 1-6, characterized in that: The system includes: a historical data collection and preprocessing module, a temperature-partial discharge characteristic correlation analysis module, a real-time demand analysis and matching module, a temperature increase strategy generation module, and a real-time data analysis and feedback module. The historical data collection and preprocessing module collects historical data on temperature and partial discharge signals from the partial discharge test. The historical data includes temperature data, environmental data, and records of partial discharge signals. The collected historical data is preprocessed and features are extracted to obtain feature parameters related to temperature changes and partial discharge characteristics, and a historical feature dataset is constructed. The temperature-partial discharge characteristic correlation analysis module aligns the temperature changes and related characteristic parameters of the historical data according to the time series based on the historical feature dataset, analyzes the correlation between temperature changes and partial discharge characteristics, and thus establishes a correlation model between temperature changes and partial discharge characteristics. The real-time demand analysis and matching module takes into account the real-time demand information of the partial discharge test, analyzes the real-time demand information in combination with historical data, and obtains historical data that matches the real-time demand information. The temperature boosting strategy generation module combines the correlation model between temperature change and partial discharge characteristics to generate a preset temperature boosting strategy. The real-time data analysis and feedback module collects real-time data from the partial discharge test, preprocesses and extracts features from the real-time data to obtain real-time characteristic parameters related to temperature changes and partial discharge characteristics; it evaluates the preset temperature increase strategy based on the real-time characteristic parameters and performs corresponding processing according to the evaluation results.

8. A temperature feeder control system based on data analysis according to claim 7, characterized in that: The historical data collection and preprocessing module includes a historical data acquisition unit, a data preprocessing unit, and a feature extraction unit. The historical data acquisition unit is responsible for collecting historical data from the partial discharge test, including temperature data, environmental data, and partial discharge signal data; the data preprocessing unit preprocesses the collected historical data; the feature extraction unit extracts features from the preprocessed data to obtain key features of temperature and partial discharge signal, forms feature vectors, and constructs a historical feature dataset. The temperature-partial discharge characteristic correlation analysis module includes a time series alignment unit, a correlation analysis and modeling unit, and a model fusion and optimization unit. The time series alignment unit performs time series alignment on the feature parameters related to temperature changes and partial discharge characteristics based on the historical feature dataset; the correlation analysis and modeling unit analyzes the correlation between temperature changes and partial discharge signal characteristics, and constructs a linear regression model and a support vector regression model based on the correlation analysis results; the model fusion and optimization unit performs weighted fusion of the prediction results of the linear regression model and the support vector regression model to form a weighted fusion model.

9. A temperature feeder control system based on data analysis according to claim 7, characterized in that: The real-time demand analysis and matching module includes a real-time demand information analysis unit and a historical data matching unit; The real-time demand information analysis unit inputs the real-time demand information of the current partial discharge test and analyzes the similarity between the real-time demand information and historical data; the historical data matching unit obtains the final matched historical data based on the analysis results of the real-time demand information analysis unit.

10. A temperature feeder control system based on data analysis according to claim 9, characterized in that: The real-time data analysis and feedback module includes a real-time data collection and analysis unit and a real-time feedback unit; The real-time data collection and analysis unit collects real-time data from the partial discharge test, performs preprocessing and feature extraction on the real-time data in the same manner as historical data, and obtains real-time feature parameters. The real-time feedback unit analyzes the real-time feature parameters and the historical feature parameters corresponding to the finally matched historical data, calculates the mean square error, and compares it with a preset threshold range. If the mean square error is within the preset threshold range, the corresponding real-time data is saved; if the mean square error exceeds the preset threshold range, a corresponding feedback mechanism is triggered to adjust the temperature increase strategy.

Images