Early warning index construction method of active transformer area state evaluation system

Abstract

The invention discloses an early warning index construction method for an active transformer area state evaluation system, and the method comprises the specific steps: collecting multi-source power generation and utilization and meteorological prediction data of an active transformer area, and obtaining a plurality of groups of sample data; establishing an active transformer area operation state evaluation index system, and inputting data into the evaluation index system; the subjective weight is determined through an analytic hierarchy process, the objective weight of the index is determined through an entropy weight method, the subjective weight and the objective weight are fused through a range maximization combination weighting method, the final weight is obtained, and an early warning index model is established; the early warning index model is put into actual operation of the transformer area, feedback data are collected, an evaluation result is calculated, and the early warning index model is optimized; and evaluating the operation state of the active transformer area power grid by using the early warning model and outputting an abnormal early warning.

Description

Technical Field

[0001] The present invention relates to the technical field of power system operation, and in particular to a method for constructing early warning indicators of an active substation status evaluation system. Background Art

[0002] Since new green energy sources such as wind power and photovoltaic power are greatly affected by weather factors, the rapid construction and grid connection of distributed power sources such as wind and solar power has greatly increased the complexity and uncertainty of the distribution system, resulting in active substations encountering problems such as photovoltaic reverse overcapacity and power return in actual management, which poses a major challenge to the safe and stable operation of the substation and the consumption of new energy. In order to accurately evaluate and warn the operating status of active substations, the quantitative indicator system results can accurately reflect the operating status and potential risks of the substation, and the real-time perception and comprehensive evaluation of the operating status of the substation can be carried out, which is conducive to flexible scheduling, selection and targeted transformation of problem substations. However, due to the short time for large-scale access to the power grid for distributed power sources, the current comprehensive evaluation and early warning research on active substations still lacks a scientific and reasonable evaluation index system and an effective empowerment method. Summary of the invention

[0003] The purpose of the present invention is to provide a method for constructing early warning indicators of an active substation status evaluation system, accurately quantify the operating status and risks of the substation, and effectively improve the safety and stability of the active substation power grid operation.

[0004] In order to achieve the above object, the present invention provides a method for constructing early warning indicators of an active area status evaluation system, and the specific steps include:

[0005] Step 1: Collect multi-source power generation and consumption and weather forecast data of active substations to obtain several groups of sample data;

[0006] Step 2, establishing an active station area operation status evaluation index system and entering data into the evaluation index system;

[0007] Step 3: Determine the subjective weight through the hierarchical analysis method, determine the objective weight of the indicator through the entropy weight method, use the range maximization combined weighting method to fuse the subjective and objective weights, obtain the final weight, and establish the early warning indicator model;

[0008] Step 4: Put the early warning indicator model into actual operation in the substation, collect feedback data, calculate evaluation results, and optimize the early warning indicator model;

[0009] Step 5: Use the early warning model to evaluate the operating status of the active substation power grid and output abnormal warnings.

[0010] Optionally, in step 3, when the range maximization combined weighting method is used to fuse the subjective and objective weights, each input data in each indicator is standardized using the upper and lower bounds of each indicator, and a standardized matrix X containing each indicator and each input data is constructed;

[0011]

[0012] In the formula, x′ nm is the standardized input data of the mth indicator in the nth group of sample data, x m is the vector of standardized input data for the mth indicator;

[0013] Set the optimal weight W and use it to calculate the evaluation result Z i ;

[0014] W=(ω1,ω2,……ω m )

[0015] Z i =(ω1x′ i1 +ω2x′ i2 +ω m x′ im )

[0016] In the formula, ω m is the weight of the mth indicator, Z i is the evaluation result of the i-th group of sample data, x im is the weight of the mth indicator in the i-th group of sample data;

[0017] Establish a combined weighted optimization model to obtain the weights of various indicators;

[0018]

[0019] In the formula, is the mean of the evaluation results, α i + is the upper bound of each indicator, α j - is the lower bound of each indicator.

[0020] Optionally, in step 3, when the objective weight of the indicator is determined by using the entropy weight method, each indicator in the evaluation indicator system in step 2 is divided into positive indicators and negative indicators, and standardized, and the information entropy E of each indicator is calculated. j ;

[0021]

[0022]

[0023] In the formula, m is the total number of sample data, n is the total number of indicators in the evaluation index system; x′tj P is the input data after annotation processing. tj P is the proportion of the input data after standardization to the total input data of the corresponding indicator. tj The larger the value, the greater the weight of the indicator in the evaluation index system;

[0024] The information entropy is used to calculate the weight of each indicator as follows:

[0025] D j =1-E j

[0026]

[0027] Where D j is the information redundancy, W j is the objective weight of the j-th indicator.

[0028] Optionally, the methods for normalizing the positive indicators and the negative indicators include:

[0029] Positive indicators:

[0030] Negative indicators:

[0031] In the formula, x tj is the input data of the jth indicator in the tth group of sample data, max(x j )、min(x j ) are the maximum and minimum values ​​of the input data in the jth indicator, respectively, x j is the average value of the data entered in the jth indicator.

[0032] Optionally, step 2 also includes: preprocessing the input data of the evaluation index system; filling in the missing values ​​in the item indicators through the Lagrange interpolation method, processing the abnormal data in the input data through the Wright test method, and processing the duplicated data in the input data through duplicate data merging.

[0033] Optionally, in step 2, during the data entry process, when there are 3 or more missing values ​​in a row for the same indicator, the indicator is deleted; when there are less than 3 missing values ​​for the same indicator, the Lagrange interpolation method is used to fill in the missing values ​​in the indicator.

[0034] Optionally, the step 2 further includes: setting a data quality threshold and calculating the quality of the input data;

[0035]

[0036] Where N satisfies |V i |<3σThe number of input data, where V i is the residual of the input data, σ is the standard deviation of the input data; M is the total number of input data in this indicator;

[0037] When the quality of input data is greater than or equal to the data quality threshold, it means that the quality of input data of this indicator is good; when the quality of input data is less than the data quality threshold, the input data is preprocessed again.

[0038] Optionally, in step 2, the extreme gradient boosting method is used to sort the indicators corresponding to the multidimensional data collected in step 1 by feature importance, so as to screen out the indicators in the evaluation indicator system.

[0039] Optionally, in step 5, the evaluation results obtained in step 4 are graded into excellent, good, medium and poor, the evaluation results are ranked, and the warning results are output;

[0040] When the evaluation result of the operating status of the target active substation is among the top 10% of the evaluation results of all active substations in the region, the early warning result is recorded as "excellent"; when the evaluation result of the operating status of the target active substation is ranked between 10% and 30% of the evaluation results of all active substations in the region, the early warning result is recorded as "good"; when the evaluation result of the operating status of the target active substation is ranked between 30% and 40% of the evaluation results of all active substations in the region, the early warning result is recorded as "medium"; when the evaluation result of the operating status of the target active substation is ranked between the bottom 60% of the evaluation results of all active substations in the region, the early warning result is recorded as "poor".

[0041] Optionally, in step 5, a warning result threshold that can trigger a warning is set, and when the warning result of the active station operation status reaches the set warning result threshold, an alarm message will be sent to the operator.

[0042] Compared with the prior art, the technical solution of the present invention has at least the following beneficial effects:

[0043] The present invention utilizes Xgboost and the correlation coefficient judgment method to optimize the index screening, obtains the active substation operation status evaluation index system, and obtains the weights of various indicators by fusing the subjective weight and the objective weight through the range maximization method, constructs an early warning model for the active substation operation status evaluation, accurately quantifies the substation operation status and risks, effectively improves the safety and stability of the active substation power grid operation, and helps the construction and development of new power systems.

[0044] The present invention integrates dimensional indicators such as photovoltaic grid-connected consumption, power quality, and power grid operation into the evaluation index system to form an early warning model for evaluating the operating status of active substations. At the same time, in view of the problems of limited active substation monitoring devices, many missing data values, and low data quality, the input data of various indicators are pre-processed to improve the data quality of the input data, obtain more accurate and effective indicator weights, improve the accuracy of substation power grid early warning, and effectively evaluate the operating status of the active substation power grid.

[0045] The present invention establishes an early warning indicator model and promptly warns operators of the results, ensuring that operators can quickly understand abnormal situations and take corresponding measures. At the same time, the early warning results are stored so that operators can predict the operating status of the substation in advance by analyzing the early warning results, and make predictions and early warnings for the substation, thereby improving the intuitiveness and convenience of the substation operation management evaluation and early warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 The present invention is a flowchart of a method for constructing early warning indicators of an active station status evaluation system.

[0047] Figure 2 This is a result diagram of the feature importance ranking of the early warning indicator construction method of the active substation status evaluation system of the present invention.

[0048] Figure 3 This is a weighting method diagram for the early warning indicator construction method of the active substation status evaluation system of the present invention. DETAILED DESCRIPTION

[0049] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. 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.

[0050] In the description of the present invention, it should be noted that the terms "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and "third" are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance.

[0051] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0052] like Figure 1 As shown, the present invention provides a method for constructing early warning indicators of an active area status evaluation system, and the specific steps are as follows:

[0053] Step 1: Collect multi-source power generation and consumption and weather forecast data in active substations.

[0054] Through public data and professional research, we use the electricity information collection system, dispatching management system, radiation illumination data platform and other channels to collect multi-dimensional data such as local photovoltaic power consumption, annual photovoltaic power generation, annual power consumption of the area, photovoltaic power conversion efficiency, user power growth rate, voltage fluctuation rate, voltage qualification rate, current fluctuation rate, distribution overload rate, etc. in the target area, and obtain several groups of sample data. Each group of sample data includes: multi-dimensional data collected every day.

[0055] Step 2: Establish an evaluation index system for the operating status of active substations.

[0056] Photovoltaic reverse overcapacity and power return are used as indicators to measure the operating status of active stations. The extreme gradient boosting (Xgboost) method is used to sort the feature importance of the indicators corresponding to the multidimensional data collected in step 1, such as Figure 2 As shown in Table 1, combined with the correlation coefficient method, data indicators with greater feature importance and higher correlation level are comprehensively selected from the aspects of photovoltaic absorption, power consumption characteristics, safe operation and power quality.

[0057] Table 1 Active station operation status evaluation indicators

[0058]

[0059] According to the active substation operation status evaluation index system in Table 1, the multidimensional data collected in step 1 is entered into the evaluation index system shown. In the process of multidimensional data entry, there are problems such as empty records or abnormal records due to abnormal operation of the data acquisition sensing device, which causes data not to be successfully recorded or recorded incorrectly, and problems such as repeated records due to repeated collection of some data during the data collection process. For the problems of empty records, abnormal records, and repeated records in the evaluation index system, it is necessary to preprocess the input data to improve the accuracy and effectiveness of the input data.

[0060] For the problem of empty records in the evaluation index system, the main method is to remove and complete the records. During the data entry process, if there are three or more missing values ​​in a row for the same index, the index will be deleted. If there are less than three missing values ​​for the same index, the Lagrange interpolation method will be used to complete the missing values ​​in the index.

[0061] The Lagrange interpolation method uses the data of the records adjacent to the missing values ​​in the same indicator to complete the missing values, and can be calculated by establishing a second-order equidistant Lagrange interpolation equation, as shown in Formula 1.

[0062]

[0063] In the formula, P2(x i ) is the formula that the missing value conforms to, x is the missing value, x i 、x k All data are entered adjacent to missing values.

[0064] In order to solve the problem of abnormal records, the Wright test method is used to process the input data. Calculate all the input data (x1, x2, x3...x n ) and some input data x i The residual V i , and calculate the standard deviation σ according to the Bessel formula, as shown in Formula 2. According to the calculated residual V i , standard deviation σ, judge the input data. If |V i |>3σ, it means that the input data is abnormal and the input data will be deleted.

[0065]

[0066] In the formula, n represents the number of input data for this indicator.

[0067] To solve the problem of duplicate records, you can merge the duplicate data.

[0068] After completing the input data preprocessing, it is necessary to check the quality of the input data and use the Wright test method to determine whether each input data in this indicator meets the requirements of |V i |<3σ. For this indicator, |V i The number of input data with |<3σ is counted and the input data quality Q is calculated, as shown in Formula 3.

[0069]

[0070] Where N satisfies |V i |<3σ, and M is the total number of input data for this indicator.

[0071] In a preferred embodiment, the data quality threshold is set to 95%. When the input data quality Q is lower than 95%, the indicator is re-preprocessed until the data quality of the indicator is greater than or equal to 95%. When the input data quality Q is greater than or equal to 95%, it means that the input data quality of the indicator is good.

[0072] Step 3: Calculate the subjective and objective weight distribution and establish an early warning indicator model.

[0073] Based on the evaluation index system for which data entry has been completed in step 2, the subjective weight of each indicator is determined by the analytic hierarchy process (AHP), and the objective weight of the indicator is determined by the entropy weight method. The subjective weight and the objective weight are integrated based on the range maximization combined weighting method to form the final weight of each indicator, and to construct an early warning index model for the operation status of the target substation power grid.

[0074] The analytic hierarchy process constructs a multi-level analysis framework consisting of a target layer, a criterion layer and a feature layer, and establishes a comparison matrix between the criterion layer and the feature layer. The elements in the comparison matrix are integers from 1 to 9 and their reciprocals, which are obtained by judging the importance of each indicator through the 1 to 9 scale system of Thomas L. Saaty. Among them, 1 means that the two indicators are equally important, 3 means that one indicator is slightly more important than another indicator, 5 means that one indicator is obviously more important than another indicator, 7 means that one indicator is very important than another indicator, and 9 means that one indicator is extremely important than another indicator; 2, 4, 6, 8 and their reciprocals indicate that the two compared indicators are between the above-mentioned importance levels.

[0075] By calculating the maximum eigenvalue λ of the comparison matrix max And consistency ratio (CI) to ensure that the consistency of the comparison matrix meets the preset standard (CI ≤ 0.1), and then determine the subjective weight of each indicator, which can be expressed as:

[0076]

[0077] In the formula, wi is the subjective weight of the ith indicator, a ij is the element in the comparison matrix, and n is the order of the matrix.

[0078] The entropy weight method determines the weight by calculating the entropy value of each factor to obtain the objective weight of each indicator.

[0079] The evaluation index system in step 2 is divided into two levels. The first level includes: photovoltaic absorption, power consumption characteristics, safe operation and power quality indicators; the second level includes: photovoltaic absorption rate, photovoltaic energy penetration rate, photovoltaic equipment utilization rate, user power growth rate, power outage operation on time restoration rate, user load growth rate, user average power outage times, distribution transformer overload rate, distribution transformer failure rate, line loss rate, voltage fluctuation rate, voltage qualification rate, current fluctuation rate, and three imbalance rates, as shown in Table 1.

[0080] The various indicators in the evaluation index system in step 2 can be divided into positive indicators and negative indicators based on factors such as power supply reliability, voltage qualification rate, power loss rate, and equipment health. At the same time, the input data in step 2 is standardized using the range method to reduce the impact of inconsistent data dimensions on the comprehensive weight results. The standardized processing methods for positive indicators and negative indicators can be expressed as follows:

[0081] Positive indicators:

[0082] Negative indicators:

[0083] In the formula, x′ tj is the input data after annotation processing, x tj is the input data of the jth indicator in the tth group of sample data, max(x j )、min(x j ) are the maximum and minimum values ​​of the input data in the jth indicator, respectively, x j is the average value of the data entered in the jth indicator.

[0084] Calculate the information entropy E of each indicator j The method is as follows:

[0085]

[0086]

[0087] In the formula, m is the total number of sample data, n is the total number of indicators in the evaluation index system; P tj P is the proportion of the input data after standardization to the total input data of the corresponding indicator. tj The larger the value, the greater the weight of the indicator in the evaluation index system.

[0088] The method of calculating the weight of each indicator using information entropy is as follows:

[0089] D j =1-E j (Formula 9)

[0090]

[0091] Where D j is the information redundancy, W j is the objective weight of the jth indicator. The greater the objective weight of the indicator, the higher the importance of the indicator in the total risk.

[0092] Based on the subjective weights of the indicators calculated by the hierarchical analysis method and the objective weights of the indicators calculated by the entropy weight method, the comprehensive weights of the indicators are calculated using the range maximization method.

[0093] The weight matrix A containing the subjective weight and objective weight of each indicator is established, which can be expressed as:

[0094]

[0095] In the formula, α zj is the weight of the jth indicator in the zth weighting method, where the weighting methods include: subjective weighting calculated by the hierarchical analysis method and objective weighting calculated by the entropy weight method. Therefore, z = 1 or 2; m is the number of indicators.

[0096] Calculate the combined weight α of each indicator zj As shown in Equation 12 and Equation 13, the upper bound α of each indicator can be obtained. i + and lower bound a j -.

[0097] α j + = max{α 1j , α 2j , ... α nj} (Formula 12)

[0098] α j - = min{α 1j , α 2j , ... α nj} (Formula 13)

[0099] Where n is the number of sample data entered into the evaluation index system obtained after preprocessing in step 2, that is, the number of entered data in each index; α nj is the weight of the jth indicator in the nth group of sample data.

[0100] Each input data in each indicator is standardized using the upper and lower bounds of each indicator, and a standardized matrix X containing each indicator and each input data is constructed, as shown in Formula 14.

[0101]

[0102] In the formula, x′ nm is the standardized input data of the mth indicator in the nth group of sample data, x m is the vector of standardized input data for the mth indicator.

[0103] The optimal weight W is set as shown in Formula 15, and the evaluation result is calculated using the optimal weight as shown in Formula 16.

[0104] W=ω1,ω2,……ω m ) (Formula 15)

[0105] l i =(ω1x′ i1 +ω2x′ i2 +ω m x′ im ) (Formula 16)

[0106] In the formula, ω m is the weight of the mth indicator, Z i is the evaluation result of the i-th group of sample data, x im is the weight of the mth indicator in the i-th group of sample data.

[0107] At this time, the variance of the evaluation results can be expressed as:

[0108]

[0109] In the formula, s 2 is the variance of the evaluation results, is the mean of the evaluation results.

[0110] A combined weighted optimization model is established, which includes: an objective function and constraints. The objective function is used to maximize the variance of the evaluation results, which can be expressed as Formula 19; the constraints are composed of a reasonable interval of each indicator and the sum of the weights of each indicator being 1, which can be expressed as Formula 20.

[0111]

[0112] At this time, the combined weighted optimization model is used to obtain the weights of each indicator in the evaluation index system and build an early warning indicator model.

[0113] Step 4: Collect feedback data and optimize the early warning indicator model.

[0114] The early warning indicator model is put into actual operation in the substation area. Feedback data is collected and stored through the dispatching management system, radiation illumination data platform, etc. The collected feedback data is scored using the weights of each indicator obtained in step 3 to obtain the evaluation results.

[0115] The collected feedback data and its evaluation results are compared, verified and analyzed with the actual operating status of the substation, and the weights of various indicators in the evaluation index system are evaluated and optimized to provide a data basis for optimizing the early warning indicator model.

[0116] Step 5: Use the early warning model to evaluate the operating status of the active substation power grid and output abnormal warnings.

[0117] The evaluation results obtained in step 4 are graded into excellent, good, medium and poor, and the evaluation results obtained in step 4 are ranked to complete the hierarchical management of the evaluation results, and the early warning results are output for subsequent analysis and processing of the operating status of the active substation.

[0118] Specifically, when the evaluation result of the operating status of the target active substation is in the top 10% of the evaluation results of all active substations in the region, the early warning result is recorded as "excellent"; when the evaluation result of the operating status of the target active substation is ranked between 10% and 30% of the evaluation results of all active substations in the region, the early warning result is recorded as "good"; when the evaluation result of the operating status of the target active substation is ranked between 30% and 40% of the evaluation results of all active substations in the region, the early warning is recorded as "medium"; when the evaluation result of the operating status of the target active substation is ranked in the bottom 60% of the evaluation results of all active substations in the region, the early warning result is recorded as "poor".

[0119] The warning result threshold that can trigger the early warning can be set according to the project needs. When the warning result of the active substation operation status reaches the set warning result threshold, an alarm message will be sent to the operator via the wireless network to ensure that the operator can quickly understand the abnormal situation and take corresponding measures; at the same time, the evaluation index system will automatically record the abnormal information, and the operator can analyze and predict the operation status of the substation through the stored abnormal information, which improves the intuitiveness and convenience of the substation operation management evaluation and early warning.

[0120] In summary, the present invention utilizes Xgboost and the correlation coefficient judgment method to optimize the index screening, obtains the active substation operation status evaluation index system, and obtains the weights of each indicator by fusing the subjective weight and the objective weight through the range maximization method, constructs an early warning model for the evaluation of the active substation operation status, accurately quantifies the substation operation status and risks, effectively improves the safety and stability of the active substation power grid operation, and helps the construction and development of new power systems.

[0121] Although the content of the present invention has been described in detail through the above preferred embodiments, it should be appreciated that the above description should not be considered as a limitation of the present invention. After reading the above content, it will be apparent to those skilled in the art that various modifications and substitutions of the present invention will occur. Therefore, the protection scope of the present invention should be limited by the appended claims.

Claims

1. A method for constructing early warning indicators for an active station status evaluation system, characterized in that: The specific steps include: Step 1: Collect multi-source power generation and consumption and weather forecast data of active substations to obtain several groups of sample data; Step 2, establishing an active station area operation status evaluation index system and entering data into the evaluation index system; Step 3: Determine the subjective weight through the hierarchical analysis method, determine the objective weight of the indicator through the entropy weight method, use the range maximization combined weighting method to fuse the subjective and objective weights, obtain the final weight, and establish the early warning indicator model; Step 4: Put the early warning indicator model into actual operation in the substation, collect feedback data, calculate evaluation results, and optimize the early warning indicator model; Step 5: Use the early warning model to evaluate the operating status of the active substation power grid and output abnormal warnings.

2. The method for constructing early warning indicators of the active area status evaluation system according to claim 1, characterized in that: In step 3, when the subjective and objective weights are integrated by using the range maximization combined weighting method, each input data in each indicator is standardized by using the upper and lower bounds of each indicator, and a standardized matrix X containing each indicator and each input data is constructed; In the formula, x′ nm is the standardized input data of the mth indicator in the nth group of sample data, x m is the vector of standardized input data for the mth indicator; Set the optimal weight W and use it to calculate the evaluation results Z i; W=(ω1,ω2,……ω) m ) Z i =(ω1x′ i1 +ω2x′ i2 +oh m x′ im ) In the formula, ω m is the weight of the mth indicator, Z i is the evaluation result of the i-th group of sample data, x im is the weight of the mth indicator in the i-th group of sample data; Establish a combined weighted optimization model to obtain the weights of various indicators; In the formula, is the mean of the evaluation results, α i + is the upper bound of each indicator, α j - is the lower bound of each indicator.

3. The method for constructing early warning indicators of the active area status evaluation system according to claim 1, characterized in that: In step 3, when the objective weight of the indicator is determined by the entropy weight method, the indicators in the evaluation indicator system in step 2 are divided into positive indicators and negative indicators, and standardized, and the information entropy E of each indicator is calculated. j ; In the formula, m is the total number of sample data, n is the total number of indicators in the evaluation index system; x′ tj P is the input data after annotation processing. tj P is the proportion of the input data after standardization to the total input data of the corresponding indicator. tj The larger the value, the greater the weight of the indicator in the evaluation index system; The information entropy is used to calculate the weight of each indicator as follows: D j =1-E i Where D j is the information redundancy, W i is the objective weight of the j-th indicator.

4. The method for constructing early warning indicators of the active area status evaluation system according to claim 3 is characterized in that: The methods for standardizing the positive indicators and negative indicators include: Positive indicators: Negative indicators: In the formula, x tj is the input data of the jth indicator in the tth group of sample data, max(x j )、min(x j ) are the maximum and minimum values ​​of the input data in the jth indicator, respectively, x j is the average value of the data entered in the jth indicator.

5. The method for constructing early warning indicators of the active area status evaluation system according to claim 1, characterized in that: The step 2 also includes: preprocessing the input data of the evaluation index system; filling in the missing values ​​in the index items by Lagrange interpolation method, processing the abnormal data in the input data by Wright test method, and processing the duplicate records in the input data by merging the duplicate data.

6. The method for constructing early warning indicators of the active area status evaluation system according to claim 5, characterized in that: In step 2, during the data entry process, when there are three or more missing values ​​in a row for the same indicator, the indicator is deleted; when there are less than three missing values ​​for the same indicator, the Lagrange interpolation method is used to fill in the missing values ​​in the indicator.

7. The method for constructing early warning indicators of the active area status evaluation system according to claim 5, characterized in that: The step 2 also includes: setting a data quality threshold and calculating the quality of the input data; Where N satisfies |V i |<3σThe number of input data, where V i is the residual of the input data, σ is the standard deviation of the input data; M is the total number of input data in this indicator; When the quality of input data is greater than or equal to the data quality threshold, it means that the quality of input data of this indicator is good; when the quality of input data is less than the data quality threshold, the input data is preprocessed again.

8. The method for constructing early warning indicators of an active area status evaluation system according to claim 1, characterized in that: In step 2, the extreme gradient boosting method is used to sort the indicators corresponding to the multidimensional data collected in step 1 according to their feature importance, and the indicators in the evaluation indicator system are screened out.

9. The method for constructing early warning indicators of an active area status evaluation system according to claim 1, characterized in that: In step 5, the evaluation results obtained in step 4 are graded into excellent, good, medium and poor, the evaluation results are ranked, and the warning results are output; When the evaluation result of the operating status of the target active substation is in the top 10% of the evaluation results of all active substations in the region, the warning result is recorded as "excellent"; When the evaluation result of the target active substation operation status ranks between 10% and 30% of the evaluation results of all active substations in the region, the warning result is recorded as "good"; When the evaluation result of the target active substation operation status ranks between 30% and 40% of the evaluation results of all active substations in the region, the warning result is recorded as "medium"; When the evaluation result of the operating status of the target active substation ranks in the bottom 60% of the evaluation results of all active substations in the region, the warning result is recorded as "poor".

10. The method for constructing early warning indicators of an active area status evaluation system according to claim 1, characterized in that: In step 5, a warning result threshold that can trigger a warning is set. When the warning result of the active station operation status reaches the set warning result threshold, an alarm message will be sent to the operator.

Images