Electric power environment positioning result optimization method based on smoothing algorithm

Abstract

The invention relates to the technical field of positioning result optimization, and discloses an electric power environment positioning result optimization method based on a smoothing algorithm, and the method comprises the optimization steps: S1, data collection: collecting original positioning data from an electric power environment positioning system; s2, data preprocessing: carrying out data cleaning and preprocessing on the original positioning data, and enhancing the data quality by using a time sequence analysis or signal processing technology; s3, bad data identification and correction: carrying out preliminary analysis on the preprocessed data, and identifying and removing data which obviously exceed a normal range or have obvious errors; s4, designing a positioning model: designing a positioning model integrated with a smoothing algorithm, wherein the positioning model is used for processing real-time data and quickly responding; s5, result evaluation: evaluating whether the positioning result after smoothing processing meets the requirement or not; and S6, optimizing and adjusting: optimizing and adjusting smoothing algorithm parameters according to an evaluation result. Positioning errors are effectively reduced, and accuracy and reliability of positioning results are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of positioning result optimization, and in particular to a method for optimizing electric power environment positioning results based on a smoothing algorithm. Background Art

[0002] With the steady growth of China's economy and the acceleration of urbanization, the demand for electricity continues to grow. This provides a broad space for development for the power industry and also promotes the continuous innovation and application of power environment positioning technology. By locating the collected location data, the power environment remote monitoring system can accurately determine the location of power facilities or fault points. This is crucial for subsequent maintenance and fault handling, and can greatly improve maintenance efficiency and accuracy.

[0003] After searching, the invention patent with Chinese patent number CN117937753A discloses a remote monitoring system for electric power environment, which belongs to the field of electric power environment operation technology. In order to solve the technical problem that the existing electric power environment monitoring is mostly to simply detect the circuit operation related conditions, and the detection and judgment of the circuit external environment is not fully ideal, which leads to unexpected fault conditions and even causes a large range of fault conditions, the system includes a data acquisition unit and a positioning unit, the data acquisition unit is connected to a data compression unit, the positioning unit is connected to the data compression unit, the data compression unit is connected to a data transmission unit, and the data transmission unit is externally connected to a data receiving and decompression unit. Compared with the prior art, the invention patent with Chinese patent number CN117937753A adopts a prediction processing unit. The prediction processing unit determines the risk factors affecting the power environment based on the acquired data, and then uses a neural network prediction model to analyze the relevant data to obtain the probability of failure in the area where the corresponding data is located. The overall data acquisition adopts a distributed design, which can obtain data from each grid area more comprehensively and accurately, so as to make advance preparations for failures in related areas and promptly repair and maintain related failures.

[0004] However, when the above-mentioned device is in use, jumping data or noise may occur in the positioning process, resulting in poor positioning results. Poor positioning results will affect the system's risk assessment of the power environment, which may cause potential safety hazards to be ignored or underestimated, increasing the possibility of accidents. Therefore, a method for optimizing power environment positioning results based on a smoothing algorithm is proposed. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art that jump data or noise may occur during the positioning process, resulting in poor positioning results. Poor positioning results will affect the system's risk assessment of the power environment, may cause potential safety hazards to be ignored or underestimated, and increase the possibility of accidents. A method for optimizing power environment positioning results based on a smoothing algorithm is proposed.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] A method for optimizing power environment positioning results based on a smoothing algorithm, the optimization steps are:

[0008] S1: Data collection: Collecting original positioning data from the power environment positioning system, including but not limited to location coordinates, timestamp and other information;

[0009] S2: Data preprocessing: Clean and preprocess the original positioning data to reduce the impact of noise and outliers, and use time series analysis or signal processing technology to enhance data quality;

[0010] S3: Bad data identification and correction: Perform preliminary analysis on preprocessed data to identify and eliminate data that is clearly outside the normal range or contains obvious errors;

[0011] S4: Positioning model design: Design a positioning model integrating a smoothing algorithm, which is used to process real-time data and respond quickly;

[0012] S5: Result evaluation: Evaluate whether the positioning result after smoothing meets the requirements, check whether the jump data or noise is eliminated, and whether the accuracy and stability of the positioning result are improved;

[0013] S6: Optimization and adjustment: According to the evaluation results, the smoothing algorithm parameters are optimized and adjusted to obtain better positioning effect.

[0014] The above further includes:

[0015] In S1, the original positioning data is collected from the power environment positioning system to ensure that the original positioning data is comprehensive and representative, including data in different time periods and different environmental conditions;

[0016] Using Kalman filtering to integrate data from different sources involves the following steps:

[0017] Initialization: Setting the initial state estimate and its covariance matrix P0;

[0018] predict:

[0019] State prediction: Use the state transfer matrix F and the optimal estimate of the previous moment To predict the current state

[0020] Covariance prediction: using the state transfer matrix F and the covariance P of the previous moment k-1 To predict the covariance P at the current moment k :P k =FP k-1 F T +Q, where Q is the covariance matrix of the process noise;

[0021] renew:

[0022] Calculate the Kalman gain: Use the prediction covariance P k and the covariance R of the observation noise to calculate the Kalman gain K k :K k =P k H T (HP k H T +R) -1 ; Where H is the observation matrix;

[0023] State update: Use Kalman gain K k , predicted value z k and predicted status To update the optimal state estimate at the current moment

[0024] Covariance update: using Kalman gain K k and the predicted covariance P k To update the covariance Pk at the current moment: Pk = (IK k H)P k ; where I is the identity matrix;

[0025] Iteration:

[0026] The optimal state estimate at the current moment The and covariance Pk are used as the initial state estimate and covariance at the next moment, and the prediction and update are repeated.

[0027] In S2, the original positioning data is cleaned and preprocessed to reduce the impact of noise and outliers, and time series analysis or signal processing techniques are used to enhance data quality:

[0028] If the data mainly exhibits time dependence (such as the trend of location coordinates changing over time), and it is necessary to predict future data values ​​or identify time patterns in the data, use time series analysis to enhance data quality; if the data is greatly affected by high-frequency noise or signal interference (such as signal attenuation, electromagnetic interference, etc.), and it is necessary to improve the signal-to-noise ratio of the data or extract useful signals, use signal processing technology to enhance data quality;

[0029] Suppose we have an electric power environment positioning system that regularly collects the location coordinates (including longitude and latitude) of the substation and related environmental parameters (such as temperature and humidity). These data are affected by many factors, including measurement errors of the equipment itself, electromagnetic interference, and weather changes, resulting in noise and outliers in the data. At the same time, these data also show certain periodic changes over time, such as changes in equipment operating status due to seasonal factors (such as temperature changes);

[0030] Then we can use signal processing techniques (such as low-pass filtering, median filtering, etc.) to remove high-frequency noise and outliers in the data. These techniques can help us smooth the data and reduce data fluctuations caused by measurement errors and electromagnetic interference. After removing the noise, we can use time series analysis techniques (such as ARI MA model, seasonal decomposition, etc.) to identify time patterns and periodic changes in the data. These techniques can help us understand the laws of data changes over time and predict future data values.

[0031] In S3, use Z-score to identify data that is clearly outside the normal range or has obvious errors. The specific steps are as follows:

[0032] Calculate the average:

[0033] Calculate the mean μ of the data set, which is the sum of all data points divided by the number of data points n:

[0034]

[0035] Among them, x i is the i-th data point in the data set, and n is the total number of data points;

[0036] Calculate the standard deviation:

[0037] Calculate the standard deviation σ of the data set, which reflects the degree of dispersion of the data points from their mean:

[0038]

[0039] Calculate the Z-score:

[0040] For each data point x in the dataset i, calculate its Z-score, which is the distance between the data point and the mean:

[0041]

[0042] Identifying outliers:

[0043] According to the calculated Z-score value, outliers are identified. If the absolute value of the Z-score is greater than a predetermined threshold (such as 3), the data point is considered to be an outlier:

[0044] |Z i >threshold

[0045] If a data point meets the above conditions, it is considered an outlier and can be further processed (such as removed, corrected, or labeled) as needed.

[0046] In S4, a positioning model integrating a smoothing algorithm is designed, and the positioning model includes a data receiving module, a smoothing algorithm selection module, an application module, a positioning result output module and a feedback module. The data receiving module is responsible for receiving data after eliminating data that is obviously beyond a normal range or has obvious errors. The data receiving module passes the received original position data to the smoothing algorithm selection module. The smoothing algorithm selection module is responsible for intelligently selecting the most suitable smoothing algorithm according to the requirements of the application scenario (such as real-time performance, positioning accuracy) and the characteristics of the received data. The smoothing algorithm selection module receives data from the data receiving module and receives optimization suggestions from the feedback module. The smoothing algorithm selection module passes the selected smoothing algorithm and its parameter configuration to the application module. The application module is responsible for applying the selected smoothing algorithm to process the pre-processed data to generate a smooth and accurate positioning trajectory. The application module receives the smoothing algorithm and its parameter configuration from the smoothing algorithm selection module, receives data from the data receiving module, the application module passes the smoothed positioning result to the positioning result output module, the positioning result output module is responsible for outputting the smoothed positioning result to the electric power environment remote monitoring system or other users in an appropriate form (such as a graphical interface, a log file, a database record, a network interface, etc.), the positioning result output module receives the smoothed positioning result from the application module, the positioning result output module outputs the positioning result to an external system or user, the feedback module is responsible for collecting feedback information (such as positioning accuracy, real-time performance, etc.) from the electric power environment remote monitoring system or other users for evaluating model performance and guiding subsequent optimization, and the feedback module passes the feedback information to the smoothing algorithm selection module so as to adjust the algorithm selection logic according to the feedback.

[0047] The specific steps of the smoothing algorithm selection module intelligently selecting the smoothing algorithm are:

[0048] Demand analysis: clarify the specific requirements of the application scenario, including real-time requirements (i.e. algorithm processing speed), positioning accuracy requirements, and restrictions on computing resources (such as CPU and memory). These requirements come from specific requirements of external systems (such as power environment remote monitoring system) or are derived by the module itself based on historical data and performance evaluation;

[0049] Data feature analysis: Perform feature analysis on the received data, including the noise level, fluctuation range, trend, periodicity, etc. The results of the feature analysis will serve as an important basis for selecting the smoothing algorithm;

[0050] Algorithm library evaluation: The smoothing algorithm selection module maintains an algorithm library, which contains different types of smoothing algorithms, such as moving average, exponential smoothing, Kalman filtering, particle filtering, etc. For each algorithm, its performance under different data characteristics is evaluated, including smoothing effect, computational complexity, real-time performance, etc.;

[0051] Algorithm selection and configuration: Based on the results of demand analysis and data feature analysis, select the smoothing algorithm that best suits the current application scenario. After selecting the algorithm, configure the algorithm parameters according to specific needs to further optimize the algorithm performance;

[0052] Performance verification: Use a portion of the test data to verify the performance of the algorithm. The verification process includes comparing the differences between the data processed by the algorithm and the original data, evaluating the real-time performance of the algorithm, and checking whether there are any abnormalities or instabilities;

[0053] Feedback and optimization: If the algorithm performs well in the performance verification, it will be used for subsequent data processing. If the performance is poor, the demand analysis, data feature analysis, and algorithm selection and configuration will be re-performed based on the feedback results.

[0054] In S5, evaluate whether the positioning result after smoothing meets the requirements, check whether the jump data or noise is eliminated, and whether the accuracy and stability of the positioning result are improved:

[0055] Correlation analysis of InSAR monitoring results with external factors includes the following steps:

[0056] Select evaluation indicators: determine the quantitative and qualitative indicators used to evaluate the performance of positioning results, for example, mean square error (MSE), root mean square error (RMSE), accuracy, recall, etc.

[0057] Qualitative evaluation: Analyze the stability and robustness of the localization model, for example, by observing how the model performs under different environmental conditions;

[0058] Quantitative evaluation: Use historical data to test the positioning model and calculate the selected quantitative indicators;

[0059] Actual environment testing: deploy the model in an actual power environment, conduct field tests, collect test data, and compare it with the actual positioning results;

[0060] Performance analysis: Analyze the results of real-world testing to evaluate the actual application performance of the model, identify the strengths and limitations of the model, and areas where improvements may be needed.

[0061] In qualitative assessment:

[0062] Visualize positioning results: Use tools such as charts or maps to visualize positioning results. For example, use a time series graph to show the change of positioning data over time, or use a scatter plot to show the positioning coordinates at different time points.

[0063] Check for jumps or noise: Observe the visualization to check for obvious jumps or noise, which may appear as sudden spikes or fluctuations in a time series plot, for example.

[0064] Analyze the stability of the model: observe the performance of the model under different environmental conditions, for example, compare the positioning accuracy and stability of the model indoors and outdoors and in different weather conditions;

[0065] Analyze the robustness of the model: evaluate the sensitivity of the model to outliers or input errors. For example, you can input some abnormal or erroneous data into the model and observe the model's response and recovery ability.

[0066] Record observations: Recording observations from the qualitative evaluation, including the model’s strengths, limitations, possible problems, etc., can help provide guidance for subsequent model improvements.

[0067] In the quantitative assessment:

[0068] Prepare test data: Select a portion of historical data as test data to ensure that the test data is representative and can fully reflect various situations that the model may encounter;

[0069] Model testing: Use test data to test the positioning model and record the model's prediction results;

[0070] Calculate evaluation indicators: Use the predicted results and actual results to calculate quantitative evaluation indicators:

[0071] Mean Square Error:

[0072]

[0073] Among them, y i is the actual value, is the predicted value, n is the number of data points;

[0074] Root Mean Square Error:

[0075]

[0076] Analyze the indicator results: Analyze the calculated indicator results to evaluate the performance of the model. For example, lower MSE and RMSE values ​​usually indicate that the model has higher prediction accuracy.

[0077] Documentation: Recording the testing process, the evaluation metrics used, and the calculated results helps ensure the reproducibility of the evaluation process and facilitates future analysis and optimization.

[0078] The present invention has the following beneficial effects:

[0079] In the present invention, by introducing a smoothing algorithm to optimize the power environment positioning result, the positioning error can be effectively reduced, the accuracy and reliability of the positioning result can be improved, and the problem of poor power environment positioning result can be avoided. BRIEF DESCRIPTION OF THE DRAWINGS

[0080] Figure 1 This is a method step diagram of a method for optimizing electric power environment positioning results based on a smoothing algorithm proposed by the present invention. DETAILED DESCRIPTION

[0081] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0082] See also Figure 1 As shown, the present invention is a method for optimizing power environment positioning results based on a smoothing algorithm, and the optimization steps are:

[0083] S1: Data collection: Collecting original positioning data from the power environment positioning system, including but not limited to location coordinates, timestamp and other information;

[0084] S2: Data preprocessing: Clean and preprocess the original positioning data to reduce the impact of noise and outliers, and use time series analysis or signal processing technology to enhance data quality;

[0085] S3: Bad data identification and correction: Perform preliminary analysis on preprocessed data to identify and eliminate data that is clearly outside the normal range or contains obvious errors;

[0086] S4: Positioning model design: Design a positioning model that integrates a smoothing algorithm. The positioning model is used to process real-time data and respond quickly.

[0087] S5: Result evaluation: Evaluate whether the positioning result after smoothing meets the requirements, check whether the jump data or noise is eliminated, and whether the accuracy and stability of the positioning result are improved;

[0088] S6: Optimization and adjustment: According to the evaluation results, the smoothing algorithm parameters are optimized and adjusted to obtain better positioning effect.

[0089] In one embodiment, for the above S1, in S1, original positioning data is collected from the power environment positioning system to ensure that the original positioning data is comprehensive and representative, including data in different time periods and under different environmental conditions;

[0090] Using Kalman filtering to integrate data from different sources involves the following steps:

[0091] Initialization: Setting the initial state estimate and its covariance matrix P0;

[0092] predict:

[0093] State prediction: Use the state transfer matrix F and the optimal estimate of the previous moment To predict the current state

[0094] Covariance prediction: using the state transfer matrix F and the covariance P of the previous moment k-1 To predict the covariance P at the current moment k :P k =FP k-1 F T +Q, where Q is the covariance matrix of the process noise;

[0095] renew:

[0096] Calculate the Kalman gain: Use the prediction covariance P k and the covariance R of the observation noise to calculate the Kalman gain K k :K k =P k H T (HP k H T +R) -1 ; Where H is the observation matrix;

[0097] State update: Use Kalman gain K k , predicted value z k and predicted status To update the optimal state estimate at the current moment

[0098] Covariance update: using Kalman gain K k and the predicted covariance P k To update the covariance Pk at the current moment: Pk = (IK k H)P k ; where I is the identity matrix;

[0099] Iteration:

[0100] The optimal state estimate at the current moment The and covariance Pk are used as the initial state estimate and covariance at the next moment, and the prediction and update are repeated.

[0101] In one embodiment, for the above S2, in S2, the original positioning data is cleaned and preprocessed to reduce the influence of noise and outliers, and time series analysis or signal processing technology is used to enhance the data quality:

[0102] If the data mainly exhibits time dependence (such as the trend of location coordinates changing over time), and it is necessary to predict future data values ​​or identify time patterns in the data, use time series analysis to enhance data quality; if the data is greatly affected by high-frequency noise or signal interference (such as signal attenuation, electromagnetic interference, etc.), and it is necessary to improve the signal-to-noise ratio of the data or extract useful signals, use signal processing technology to enhance data quality;

[0103] Suppose we have an electric power environment positioning system that regularly collects the location coordinates (including longitude and latitude) of the substation and related environmental parameters (such as temperature and humidity). These data are affected by many factors, including measurement errors of the equipment itself, electromagnetic interference, and weather changes, resulting in noise and outliers in the data. At the same time, these data also show certain periodic changes over time, such as changes in equipment operating status due to seasonal factors (such as temperature changes);

[0104] Then we can use signal processing techniques (such as low-pass filtering, median filtering, etc.) to remove high-frequency noise and outliers in the data. These techniques can help us smooth the data and reduce data fluctuations caused by measurement errors and electromagnetic interference. After removing the noise, we can use time series analysis techniques (such as ARI MA model, seasonal decomposition, etc.) to identify time patterns and periodic changes in the data. These techniques can help us understand the laws of data changes over time and predict future data values.

[0105] In one embodiment, for the above S3, in S3, Z-score is used to identify data that is obviously beyond the normal range or has obvious errors, and the specific steps are:

[0106] Calculate the average:

[0107] Calculate the mean μ of the data set, which is the sum of all data points divided by the number of data points n:

[0108]

[0109] Among them, x i is the i-th data point in the data set, and n is the total number of data points;

[0110] Calculate the standard deviation:

[0111] Calculate the standard deviation σ of the data set, which reflects the degree of dispersion of the data points from their mean:

[0112]

[0113] Calculate the Z-score:

[0114] For each data point x in the dataset i , calculate its Z-score, which is the distance between the data point and the mean:

[0115]

[0116] Identifying outliers:

[0117] According to the calculated Z-score value, outliers are identified. If the absolute value of the Z-score is greater than a predetermined threshold (such as 3), the data point is considered to be an outlier:

[0118] |Z i >threshold

[0119] If a data point meets the above conditions, it is considered an outlier and can be further processed (such as removed, corrected, or labeled) as needed.

[0120] In one embodiment, for the above S4, in S4, a positioning model integrating a smoothing algorithm is designed, and the positioning model includes a data receiving module, a smoothing algorithm selection module, an application module, a positioning result output module and a feedback module. The data receiving module is responsible for receiving the data after eliminating the data that obviously exceeds the normal range or has obvious errors. The data receiving module passes the received original position data to the smoothing algorithm selection module. The smoothing algorithm selection module is responsible for intelligently selecting the most suitable smoothing algorithm according to the requirements of the application scenario (such as real-time performance, positioning accuracy) and the characteristics of the received data. The smoothing algorithm selection module receives data from the data receiving module and receives optimization suggestions from the feedback module. The smoothing algorithm selection module passes the selected smoothing algorithm and its parameter configuration to the application module. The application module is responsible for applying the selected smoothing algorithm to process the pre-processed data to generate a smoothed image. , accurate positioning trajectory, the application module receives the smoothing algorithm and its parameter configuration from the smoothing algorithm selection module, receives data from the data receiving module, the application module passes the smoothed positioning result to the positioning result output module, the positioning result output module is responsible for outputting the smoothed positioning result in an appropriate form (such as graphical interface, log file, database record, network interface, etc.) to the electric power environment remote monitoring system or other users, the positioning result output module receives the smoothed positioning result from the application module, the positioning result output module outputs the positioning result to the external system or user, the feedback module is responsible for collecting feedback information (such as positioning accuracy, real-time performance, etc.) from the electric power environment remote monitoring system or other users, which is used to evaluate the model performance and guide subsequent optimization, the feedback module passes the feedback information to the smoothing algorithm selection module so as to adjust the algorithm selection logic according to the feedback.

[0121] In one embodiment, for the above-mentioned smoothing algorithm selection module, the specific steps of the smoothing algorithm selection module intelligently selecting a smoothing algorithm are:

[0122] Demand analysis: clarify the specific requirements of the application scenario, including real-time requirements (i.e. algorithm processing speed), positioning accuracy requirements, and restrictions on computing resources (such as CPU and memory). These requirements come from specific requirements of external systems (such as power environment remote monitoring system) or are derived by the module itself based on historical data and performance evaluation;

[0123] Data feature analysis: Perform feature analysis on the received data, including the noise level, fluctuation range, trend, periodicity, etc. The results of feature analysis will serve as an important basis for selecting smoothing algorithms;

[0124] Algorithm library evaluation: The smoothing algorithm selection module maintains an algorithm library, which contains different types of smoothing algorithms, such as moving average, exponential smoothing, Kalman filtering, particle filtering, etc. For each algorithm, its performance under different data characteristics is evaluated, including smoothing effect, computational complexity, real-time performance, etc.

[0125] Algorithm selection and configuration: Based on the results of demand analysis and data feature analysis, select the smoothing algorithm that best suits the current application scenario. After selecting the algorithm, configure the algorithm parameters according to specific needs to further optimize the algorithm performance;

[0126] Performance verification: Use a portion of the test data to verify the performance of the algorithm. The verification process includes comparing the differences between the data processed by the algorithm and the original data, evaluating the real-time performance of the algorithm, and checking whether there are any abnormalities or instabilities;

[0127] Feedback and optimization: If the algorithm performs well in the performance verification, it will be used for subsequent data processing. If the performance is poor, the demand analysis, data feature analysis, and algorithm selection and configuration will be re-performed based on the feedback results.

[0128] In one embodiment, for the above S5, in S5, it is evaluated whether the positioning result after smoothing meets the requirements, whether the jump data or noise is eliminated, and whether the accuracy and stability of the positioning result are improved:

[0129] Correlation analysis of InSAR monitoring results with external factors includes the following steps:

[0130] Select evaluation indicators: determine the quantitative and qualitative indicators used to evaluate the performance of positioning results, for example, mean square error (MSE), root mean square error (RMSE), accuracy, recall, etc.

[0131] Qualitative evaluation: Analyze the stability and robustness of the localization model, for example, by observing how the model performs under different environmental conditions;

[0132] Quantitative evaluation: Use historical data to test the positioning model and calculate the selected quantitative indicators;

[0133] Actual environment testing: deploy the model in an actual power environment, conduct field tests, collect test data, and compare it with the actual positioning results;

[0134] Performance analysis: Analyze the results of real-world testing to evaluate the actual application performance of the model, identify the strengths and limitations of the model, and areas where improvements may be needed.

[0135] In one embodiment, for the above qualitative evaluation, in the qualitative evaluation:

[0136] Visualize positioning results: Use tools such as charts or maps to visualize positioning results. For example, use a time series graph to show the change of positioning data over time, or use a scatter plot to show the positioning coordinates at different time points.

[0137] Check for jumps or noise: Observe the visualization to check for obvious jumps or noise, which may appear as sudden spikes or fluctuations in a time series plot, for example.

[0138] Analyze the stability of the model: observe the performance of the model under different environmental conditions, for example, compare the positioning accuracy and stability of the model indoors and outdoors and in different weather conditions;

[0139] Analyze the robustness of the model: evaluate the sensitivity of the model to outliers or input errors. For example, you can input some abnormal or erroneous data into the model and observe the model's response and recovery ability.

[0140] Record observations: Recording observations from the qualitative evaluation, including the model’s strengths, limitations, possible problems, etc., can help provide guidance for subsequent model improvements.

[0141] In one embodiment, for the above quantitative evaluation, in the quantitative evaluation:

[0142] Prepare test data: Select a portion of historical data as test data to ensure that the test data is representative and can fully reflect various situations that the model may encounter;

[0143] Model testing: Use test data to test the positioning model and record the model's prediction results;

[0144] Calculate evaluation indicators: Use the predicted results and actual results to calculate quantitative evaluation indicators:

[0145] Mean Square Error:

[0146]

[0147] Among them, y i is the actual value, is the predicted value, n is the number of data points;

[0148] Root Mean Square Error:

[0149]

[0150] Analyze the indicator results: Analyze the calculated indicator results to evaluate the performance of the model. For example, lower MSE and RMSE values ​​usually indicate that the model has higher prediction accuracy.

[0151] Documentation: Recording the testing process, the evaluation metrics used, and the calculated results helps ensure the reproducibility of the evaluation process and facilitates future analysis and optimization.

[0152] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for optimizing power environment positioning results based on a smoothing algorithm, characterized in that: The optimization steps are: S1: Data collection: Collect raw positioning data from the power environment positioning system; S2: Data preprocessing: Clean and preprocess the original positioning data, and use time series analysis or signal processing technology to enhance data quality; S3: Bad data identification and correction: Perform preliminary analysis on preprocessed data to identify and eliminate data that is clearly outside the normal range or contains obvious errors; S4: Positioning model design: Design a positioning model integrating a smoothing algorithm, which is used to process real-time data and respond quickly; S5: Result evaluation: evaluate whether the positioning result after smoothing meets the requirements; S6: Optimization and adjustment: According to the evaluation results, the smoothing algorithm parameters are optimized and adjusted.

2. According to the method for optimizing the electric power environment positioning result based on the smoothing algorithm of claim 1, it is characterized in that: In S1, the original positioning data is collected from the power environment positioning system to ensure that the original positioning data is comprehensive and representative, including data in different time periods and different environmental conditions; Using Kalman filtering to integrate data from different sources involves the following steps: Initialization: Setting the initial state estimate and its covariance matrix P0; predict: State prediction: Use the state transfer matrix F and the optimal estimate of the previous moment To predict the current state Covariance prediction: using the state transfer matrix F and the covariance P of the previous moment k-1 To predict the covariance P at the current moment k :P k =FP k-1 F T +Q, where Q is the covariance matrix of the process noise; renew: Calculate the Kalman gain: Use the prediction covariance P k and the covariance R of the observation noise to calculate the Kalman gain K k :K k =P k H T (HP k H T +R) -1 ; Where H is the observation matrix; State update: Use Kalman gain K k , predicted value z k and predicted status To update the optimal state estimate at the current moment Covariance update: using Kalman gain K k and the predicted covariance P k To update the covariance Pk at the current moment: Pk = (IK k H)P k ; where I is the identity matrix; Iteration: The optimal state estimate at the current moment The and covariance Pk are used as the initial state estimate and covariance at the next moment, and the prediction and update are repeated.

3. The method for optimizing electric power environment positioning results based on a smoothing algorithm according to claim 1, characterized in that: In S2, the original positioning data is cleaned and preprocessed, and time series analysis or signal processing techniques are used to enhance data quality: If the data mainly exhibits time dependence and it is necessary to predict future data values ​​or identify time patterns in the data, use time series analysis to enhance data quality. If the data is greatly affected by high-frequency noise or signal interference and it is necessary to improve the signal-to-noise ratio of the data or extract useful signals, use signal processing technology to enhance data quality.

4. The method for optimizing electric power environment positioning results based on a smoothing algorithm according to claim 1, characterized in that: In S3, use Z-score to identify data that is clearly outside the normal range or has obvious errors. The specific steps are as follows: Calculate the average: Calculate the mean μ of the data set, which is the sum of all data points divided by the number of data points n: Among them, x i is the i-th data point in the data set, and n is the total number of data points; Calculate the standard deviation: Calculate the standard deviation σ of the data set, which reflects the degree of dispersion of the data points from their mean: Calculate the Z-score: For each data point x in the dataset i , calculate its Z-score, which is the distance between the data point and the mean: Identifying outliers: According to the calculated Z-score value, outliers are identified. If the absolute value of the Z-score is greater than a predetermined threshold, the data point is considered to be an outlier: |Z i >threshold If a data point satisfies the above conditions, it is considered an outlier.

5. The method for optimizing electric power environment positioning results based on a smoothing algorithm according to claim 1, characterized in that: In S4, a positioning model integrating a smoothing algorithm is designed, and the positioning model includes a data receiving module, a smoothing algorithm selection module, an application module, a positioning result output module and a feedback module. The data receiving module is responsible for receiving the data after eliminating the data that obviously exceeds the normal range or has obvious errors. The data receiving module passes the received original position data to the smoothing algorithm selection module. The smoothing algorithm selection module is responsible for intelligently selecting a smoothing algorithm according to the requirements of the application scenario and the characteristics of the received data. The smoothing algorithm selection module receives data from the data receiving module and receives optimization suggestions from the feedback module. The smoothing algorithm selection module passes the selected smoothing algorithm and its parameter configuration to the application module. The application module is responsible for applying the selected smoothing algorithm to the preprocessed data. Processing is performed to generate a positioning trajectory. The application module receives the smoothing algorithm and its parameter configuration from the smoothing algorithm selection module and receives data from the data receiving module. The application module passes the positioning result after smoothing to the positioning result output module. The positioning result output module is responsible for outputting the positioning result after smoothing to the electric power environment remote monitoring system or other users. The positioning result output module receives the positioning result after smoothing from the application module and outputs the positioning result to an external system or user. The feedback module is responsible for collecting feedback information from the electric power environment remote monitoring system or other users for evaluating model performance and guiding subsequent optimization. The feedback module passes the feedback information to the smoothing algorithm selection module so as to adjust the algorithm selection logic according to the feedback.

6. The method for optimizing electric power environment positioning results based on a smoothing algorithm according to claim 5 is characterized in that: The specific steps of the smoothing algorithm selection module intelligently selecting the smoothing algorithm are: Demand analysis: clarify the specific requirements of the application scenario, including real-time requirements, positioning accuracy requirements, and limitations on computing resources; Data feature analysis: Perform feature analysis on the received data. The result of the feature analysis will serve as an important basis for selecting the smoothing algorithm. Algorithm library evaluation: The smoothing algorithm selection module maintains an algorithm library, which contains a variety of different types of smoothing algorithms. For each algorithm, its performance under different data characteristics is evaluated; Algorithm selection and configuration: Based on the results of demand analysis and data feature analysis, select the smoothing algorithm that best suits the current application scenario. After selecting the algorithm, configure the algorithm parameters according to specific needs; Performance verification: Use a portion of the test data to verify the performance of the algorithm. The verification process includes comparing the differences between the data processed by the algorithm and the original data, evaluating the real-time performance of the algorithm, and checking whether there are any abnormalities or instabilities; Feedback and optimization: If the algorithm performs well in the performance verification, it will be used for subsequent data processing. If the performance is poor, the demand analysis, data feature analysis, and algorithm selection and configuration will be re-performed based on the feedback results.

7. The method for optimizing electric power environment positioning results based on a smoothing algorithm according to claim 1, characterized in that: In S5, it is evaluated whether the positioning result after smoothing meets the requirements: Correlation analysis between InSAR monitoring results and external factors includes the following steps: Select evaluation indicators: determine the quantitative and qualitative indicators used to evaluate the performance of positioning results; Qualitative evaluation: Analyze the stability and robustness of the positioning model; Quantitative evaluation: Use historical data to test the positioning model and calculate the selected quantitative indicators; Actual environment testing: deploy the model in an actual power environment, conduct field tests, collect test data, and compare it with the actual positioning results; Performance analysis: Analyze the results of actual environment tests and evaluate the actual application performance of the model.

8. The method for optimizing electric power environment positioning results based on a smoothing algorithm according to claim 7, characterized in that: In qualitative assessment: Visual positioning results: Visual positioning results; Check for jumps or noise: Observe the visualization results to check if there are obvious jumps or noise. Analyze the stability of the model: observe the performance of the model under different environmental conditions; Analyze the robustness of the model: assess how sensitive the model is to outliers or input errors; Record Observations: Record observations from qualitative assessments.

9. The method for optimizing electric power environment positioning results based on a smoothing algorithm according to claim 7, characterized in that: In the quantitative assessment: Prepare test data: select a part of historical data as test data; Model testing: Use test data to test the positioning model and record the model's prediction results; Calculate evaluation indicators: Use the predicted results and actual results to calculate quantitative evaluation indicators: Mean Square Error: Among them, y i is the actual value, is the predicted value, n is the number of data points; Root Mean Square Error: Analyze indicator results: Analyze the calculated indicator results and evaluate the performance of the model; Documentation: Record the testing process, the evaluation metrics used, and the calculated results.

Images