Concentrator with power-off intelligent recovery function

Abstract

The invention relates to the technical field of concentrators, in particular to a concentrator with a power-off intelligent recovery function, which comprises a data acquisition module, a power-off intelligent recovery module and a power-off intelligent recovery module, and is characterized in that the data acquisition module is used for acquiring a de-noising neighborhood electric energy data point sequence corresponding to a power-off time stage of the concentrator; the prediction weight acquisition module is used for acquiring a prediction weight value of the de-noised left neighborhood electric energy data point sequence and a prediction weight value of the de-noised right neighborhood electric energy data point sequence; and the prediction module is used for obtaining target prediction electric energy data in the power-off time stage according to the prediction weight values of the de-noised left neighborhood electric energy data point sequence and the de-noised right neighborhood electric energy data point sequence. According to the method, the accuracy of electric energy prediction in the power-off time stage can be improved, so that the accuracy of electric energy prediction interpolation in the power-off time stage can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of concentrators, and in particular to a concentrator with a power failure intelligent recovery function. Background Art

[0002] A concentrator is a device for centralized collection and communication management of electric energy data, installed at the lower end of a distribution transformer or at the low-voltage incoming line of a building. It serves as a bridge between the smart meter collection terminal and the master station system. Currently, due to environmental factors and faults in the concentrator itself, the concentrator may experience power outages, resulting in loss of data collected by the concentrator during the power outage. To ensure data integrity, maintain system coordination, and avoid misjudgments and misoperations, it is usually necessary to interpolate the electric energy during the power outage. Currently, an ARIMA forecasting model is usually used to predict the electric energy during the power outage, and then the predicted data is interpolated to the corresponding position. The data used for prediction is generally data before the power outage. However, this existing forecasting and interpolation method not only contains noisy data in the data used for prediction, but also has a low accuracy of the prediction results obtained based on the data before the power outage, resulting in low accuracy or reliability of the predicted and interpolated data during the power outage. Therefore, how to improve the accuracy or reliability of forecasting and interpolation has become an urgent problem that needs to be solved. Summary of the Invention

[0003] In order to solve the above problems, the present invention provides a concentrator with a power failure intelligent recovery function. The technical solutions adopted are as follows: An embodiment of the present invention provides a concentrator with a power outage intelligent recovery function, the concentrator with a power outage intelligent recovery function includes: A data acquisition module is used to acquire a denoised neighborhood electric energy data point sequence corresponding to the power outage time stage of the concentrator, wherein the denoised neighborhood electric energy data point sequence includes a denoised left neighborhood electric energy data point sequence and a denoised right neighborhood electric energy data point sequence; a prediction weight acquisition module, configured to obtain the prediction credibility of the denoised neighborhood electric energy data point sequence based on the trend component data point sequence and the periodic component data point sequence obtained by time series decomposition of the denoised neighborhood electric energy data point sequence, and the peak points on the frequency domain signal obtained by frequency domain conversion of the periodic component data point sequence, and to obtain the prediction weight value of the denoised left neighborhood electric energy data point sequence and the prediction weight value of the denoised right neighborhood electric energy data point sequence based on the prediction credibility of the denoised left neighborhood electric energy data point sequence and the denoised right neighborhood electric energy data point sequence; The prediction module is used to obtain the target predicted electric energy data in the power outage time stage according to the prediction weight values ​​of the denoised left neighborhood electric energy data point sequence and the denoised right neighborhood electric energy data point sequence.

[0004] Beneficial effects: The present invention includes a data acquisition module for acquiring a denoised neighborhood electric energy data point sequence corresponding to the power outage time stage of the concentrator; a prediction weight acquisition module for obtaining the prediction credibility of the denoised neighborhood electric energy data point sequence based on the trend component data point sequence and the periodic component data point sequence obtained by time series decomposition of the denoised neighborhood electric energy data point sequence and the peak points on the frequency domain signal obtained by frequency domain conversion of the periodic component data point sequence, and obtaining the prediction weight value of the denoised left neighborhood electric energy data point sequence and the prediction weight value of the denoised right neighborhood electric energy data point sequence based on the prediction credibility of the denoised left neighborhood electric energy data point sequence and the denoised right neighborhood electric energy data point sequence; a prediction module for obtaining the target predicted electric energy data in the power outage time stage based on the prediction weight values ​​of the denoised left neighborhood electric energy data point sequence and the denoised right neighborhood electric energy data point sequence. Moreover, the present invention can improve the accuracy of electric energy prediction during the power outage time period based on the denoised left neighboring electric energy data point sequence and the denoised right neighboring electric energy data point sequence and the prediction weight values ​​of the denoised left neighboring electric energy data point sequence and the denoised right neighboring electric energy data point sequence, thereby improving the accuracy of electric energy prediction interpolation during the power outage time period. BRIEF DESCRIPTION OF THE DRAWINGS

[0005] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0006] Figure 1 This is a structural block diagram of a concentrator with a power outage intelligent recovery function according to the present invention. DETAILED DESCRIPTION

[0007] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field fall within the scope of protection of the embodiments of the present invention.

[0008] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0009] This embodiment provides a concentrator with a power outage intelligent recovery function, which is described in detail as follows: like Figure 1 As shown, this embodiment provides a concentrator with a power outage intelligent recovery function, including: The data acquisition module 01 is used to obtain a sequence of denoised neighborhood electric energy data points corresponding to the power outage time stage of the concentrator.

[0010] The purpose of this embodiment is to improve the accuracy of the ARIMA model in predicting the electric energy during the power outage period of the concentrator, and this embodiment mainly predicts the electric energy during the power outage period by denoising the data used for prediction and then analyzing the weights of the data before and after the power outage for the prediction after denoising. That is, after denoising, the electric energy during the power outage period is predicted by a two-way prediction method, thereby improving the accuracy of the electric energy prediction during the power outage period, thereby improving the accuracy and reliability of the electric energy prediction interpolation during the power outage period. Since the power prediction method for each concentrator in the power outage time stage is the same in this embodiment, and for the sake of ease of understanding, this embodiment will subsequently describe the power prediction process of any power outage time stage as an example, that is, the subsequent power outage time stages are all the same power outage time stages, and the subsequent concentrators are the same concentrator; in addition, predicting the power in the power outage time stage is essentially predicting the power metering data actually generated by the power meter in the power outage time stage, and the concentrator generally manages multiple power meters, so it is necessary to predict the power metering data actually generated by all power meters managed by the concentrator in the power outage time stage, and since the process of predicting the power metering data actually generated by different power meters in the power outage time stage is consistent in this embodiment, in order to facilitate understanding, this embodiment will subsequently describe the process of predicting the power metering data actually generated by any power meter Q managed by the concentrator in the power outage time stage as an example, and the power metering data will be collectively referred to as power data, that is, the power data used in the subsequent prediction of this embodiment and the predicted power data all belong to the power meter Q.

[0011] In this embodiment, the data before and after the power outage are needed in the process of predicting the electric energy data of the power outage time stage. Therefore, this embodiment needs to first obtain the left neighboring electric energy data point sequence corresponding to the left neighboring time stage of the power outage time stage and the right neighboring electric energy data point sequence corresponding to the right neighboring time stage. Then the specific acquisition process of the left neighboring electric energy data point sequence corresponding to the left neighboring time stage of the power outage time stage and the right neighboring electric energy data point sequence corresponding to the right neighboring time stage is as follows: First, the left neighboring time stage and the right neighboring time stage of the power outage time stage are obtained, and the left neighboring time stage is located in front of the power outage time stage in time and adjacent to the power outage time stage, and the right neighboring time stage is located in the back of the power outage time stage in time and adjacent to the power outage time stage; in addition, this embodiment sets the length of the left neighboring time stage and the right neighboring time stage according to the data length frequently used when the ARIMA model is used for prediction. Since in general, when the ARIMA model is used for prediction, the data length used for prediction is generally 3 to 5 times the length of the predicted data, if this embodiment can select the data length used for prediction to be 5 times the length of the predicted data, then the time lengths of the left neighboring time stage and the right neighboring time stage in this embodiment are respectively 5 times the length of the power outage time stage.

[0012] After obtaining the left neighboring time stage and the right neighboring time stage of the power outage time stage, all the electric energy data of the electric energy meter Q collected by the concentrator in the left neighboring time stage are obtained, and all the electric energy data of the electric energy meter Q collected by the concentrator in the left neighboring time stage are arranged in the order of collection, and the arrangement result is recorded as the left neighboring electric energy sequence corresponding to the left neighboring time stage, and all the electric energy data of the electric energy meter Q collected by the concentrator in the right neighboring time stage are obtained, and all the electric energy data of the electric energy meter Q collected by the concentrator in the right neighboring time stage are arranged in the order of collection, and the arrangement result is recorded as the right neighboring electric energy sequence corresponding to the right neighboring time stage; then a two-dimensional coordinate system is constructed, and the horizontal axis of the two-dimensional coordinate system represents time, The vertical axis represents the electric energy data, and then each electric energy data in the left neighborhood electric energy sequence and the right neighborhood electric energy sequence and the time when the corresponding electric energy data is collected are mapped to the two-dimensional coordinate system to obtain the electric energy data points corresponding to each electric energy data in the left neighborhood electric energy sequence and the electric energy data points corresponding to each electric energy data in the right neighborhood electric energy sequence. The time series sequence constructed by the electric energy data points corresponding to all the electric energy data in the left neighborhood electric energy sequence is recorded as the left neighborhood electric energy data point sequence corresponding to the left neighborhood time stage, and the time series sequence constructed by the electric energy data points corresponding to all the electric energy data in the right neighborhood electric energy sequence is recorded as the right neighborhood electric energy data point sequence corresponding to the right neighborhood time stage. The left neighborhood electric energy data point sequence and the right neighborhood electric energy data point sequence both belong to the neighborhood electric energy data point sequence.

[0013] Since the concentrator generates noise data during the data collection process, and the presence of noise data will affect the prediction accuracy, denoising processing is required before making a prediction to obtain a denoised neighborhood electric energy data point sequence corresponding to the power outage time stage, thereby eliminating the interference of noise on the prediction; and since compared with the non-noise data points collected by the conventional concentrator, the noise data points usually have sudden jump or outlier characteristics. Sudden jump refers to a large difference between the data point and the adjacent data points, and outlier refers to a large difference in the degree of change with the nearby data. Therefore, this embodiment will next analyze the sudden jump or outlier characteristics of the electric energy data points in the neighborhood electric energy data point sequence to screen out the noise data points in the neighborhood electric energy data point sequence, and perform noise removal to obtain a denoised neighborhood electric energy data point sequence. The specific process is as follows: First, according to the distance between the electric energy data point in the left neighborhood electric energy data point sequence and the adjacent electric energy data point of the corresponding electric energy data point, the noise data point in the left neighborhood electric energy data point sequence is obtained, and the noise data point in the left neighborhood electric energy data point sequence is interpolated and replaced, and the sequence after the noise data point is replaced is recorded as the denoised left neighborhood electric energy data point sequence of the left neighborhood electric energy data point sequence; and the denoised right neighborhood electric energy data point sequence of the right neighborhood electric energy data point sequence and the denoised left neighborhood electric energy data point sequence of the left neighborhood electric energy data point sequence are obtained. The method is the same, so the acquisition process of the denoised right neighborhood electric energy data point sequence will not be described later. The denoised left neighborhood electric energy data point sequence and the denoised right neighborhood electric energy data point sequence both belong to the denoised neighborhood electric energy data point sequence corresponding to the power-off time stage of the concentrator; in this embodiment, the implementer can choose a method for interpolating and replacing the noise data points in the left neighborhood electric energy data point sequence according to actual conditions, such as selecting linear interpolation or moving average method to interpolate and replace the noise data points in the neighborhood electric energy data point sequence.

[0014] In this embodiment, the specific process of obtaining the noise data point in the left neighboring electric energy data point sequence based on the distance between the electric energy data point in the left neighboring electric energy data point sequence and the adjacent electric energy data point of the corresponding electric energy data point is: first, based on the adjacent electric energy data points of each electric energy data point in the left neighboring electric energy data point sequence, the neighborhood electric energy distance of the electric energy data point in the left neighboring electric energy data point sequence is obtained; and then, based on the neighborhood electric energy distance of each electric energy data point in the left neighboring electric energy data point sequence, the noise index value of the electric energy data point in the left neighboring electric energy data point sequence is obtained; and then, the noise data point in the left neighboring electric energy data point sequence is obtained based on the noise index value.

[0015] In this embodiment, the specific process of obtaining the neighborhood electric energy distance of the electric energy data point in the left neighborhood electric energy data point sequence is as follows: for the g-th electric energy data point in the left neighborhood electric energy data point sequence, if g is 1, the Euclidean distance between the g-th electric energy data point and the g+1-th electric energy data point in the left neighborhood electric energy data point sequence is used as the neighborhood electric energy distance of the g-th electric energy data point; if g is A, the Euclidean distance between the g-1-th electric energy data point and the g-th electric energy data point in the left neighborhood electric energy data point sequence is used as the neighborhood electric energy distance of the g-th electric energy data point; if g is not equal to 1 and A, the g-th electric energy data point in the left neighborhood electric energy data point sequence is used as the neighborhood electric energy distance of the g-th electric energy data point. The Euclidean distance between the g-1th electric energy data point and the g+1th electric energy data point in the left neighborhood electric energy data point sequence is added to the Euclidean distance between the gth electric energy data point and the g+1th electric energy data point in the left neighborhood electric energy data point sequence, which is the neighborhood electric energy distance of the gth electric energy data point. That is, if g is not equal to 1 and A, then the sum of D1 and D2 is the neighborhood electric energy distance of the gth electric energy data point, A is the total number of electric energy data points in the left neighborhood electric energy data point sequence, D1 is the Euclidean distance between the gth electric energy data point and the g-1th electric energy data point in the left neighborhood electric energy data point sequence, and D2 is the Euclidean distance between the gth electric energy data point and the g+1th electric energy data point in the left neighborhood electric energy data point sequence.

[0016] In this embodiment, according to the neighborhood electric energy distance of each electric energy data point in the left neighborhood electric energy data point sequence, the specific process of obtaining the noise index value of the electric energy data point in the left neighborhood electric energy data point sequence is as follows: for the a-th electric energy data point in the left neighborhood electric energy data point sequence, a is greater than 1 and less than A: first calculate the absolute value of the difference between the neighborhood electric energy distance of the a-1-th electric energy data point in the left neighborhood electric energy data point sequence and the neighborhood electric energy distance of the a-th electric energy data point, and record it as the first difference value; calculate the neighborhood electric energy distance of the a+1-th electric energy data point in the left neighborhood electric energy data point sequence and the absolute value of the difference between the neighborhood electric energy distance of the a- ... The absolute value of the difference between the neighborhood electric energy distances of the a-th electric energy data point is recorded as the second difference value, and then the sum of the first difference value and the second difference value is calculated, and then the sum of the first difference value and the second difference value is normalized, and the result of the normalization is recorded as the comprehensive difference value, the neighborhood electric energy distance of the a-th electric energy data point is normalized, and the result of the normalization is recorded as the normalized neighborhood electric energy distance, and the sum of the normalized neighborhood electric energy distance and the comprehensive difference is calculated and used as the noise index value of the a-th electric energy data point; and the specific calculation expression of the noise index value of the a-th electric energy data point is:

[0017] in, is the noise index value of the a-th electric energy data point, The value of is 0 to 1, Norm() is the normalization function, is the neighborhood electric energy distance of the ath electric energy data point in the left neighborhood electric energy data point sequence, is the neighborhood electric energy distance of the a-1th electric energy data point in the left neighborhood electric energy data point sequence, is the neighborhood electric energy distance of the a+1th electric energy data point in the left neighborhood electric energy data point sequence. When the value is larger, it indicates that the sudden jump feature of the a-th electric energy data point is more obvious. When the value is larger, the outlier characteristics of the a-th electric energy data point are more obvious, and when and The bigger it is, The larger the The larger the value is, the more likely the a-th electric energy data point is noise, and vice versa. In addition, this embodiment does not calculate the noise index value and judge the noise for the first and last electric energy data points in the left neighborhood electric energy data point sequence.

[0018] In this embodiment, the specific process of obtaining the noise data points in the left neighborhood electric energy data point sequence according to the noise index value is: for the ath electric energy data point in the left neighborhood electric energy data point sequence, determine whether the noise index value of the ath electric energy data point is greater than the preset noise threshold; if so, determine that the ath electric energy data point is a noise data point, and record the ath electric energy data point as a noise data point; otherwise, determine that the ath electric energy data point is not a noise data point; and in specific applications, the implementer can set the preset noise threshold according to actual conditions, experimental statistics and the value range of the noise index value. For example, in the above-mentioned embodiment, through normalization, the data points with noise characteristics and the data points with normal fluctuation characteristics are divided at the two ends of the noise index value range. In order to ensure the reliability of noise data point screening, this embodiment can set the preset noise threshold to 0.5.

[0019] Therefore, this embodiment can obtain the denoised left-neighborhood electric energy data point sequence and the denoised right-neighborhood electric energy data point sequence corresponding to the power-off time stage of the concentrator through the above process.

[0020] The prediction weight acquisition module 02 is used to obtain the prediction credibility of the denoised neighborhood electric energy data point sequence based on the trend component data point sequence and the periodic component data point sequence obtained by time series decomposition of the denoised neighborhood electric energy data point sequence and the peak points on the frequency domain signal obtained by frequency domain conversion of the periodic component data point sequence, and to obtain the prediction weight value of the denoised left neighborhood electric energy data point sequence and the prediction weight value of the denoised right neighborhood electric energy data point sequence based on the prediction credibility of the denoised left neighborhood electric energy data point sequence and the denoised right neighborhood electric energy data point sequence.

[0021] In order to improve the accuracy of the prediction, this embodiment chooses to adopt a bidirectional prediction method to predict the electric energy data in the power outage time stage, that is, this embodiment predicts the electric energy data in the power outage time stage from two directions: before and after the power outage. Since the credibility of the prediction results obtained by predicting the data before the power outage and the data after the power outage is different, this embodiment will then need to analyze the denoised left neighborhood electric energy data point sequence and the denoised right neighborhood electric energy data point sequence, and obtain the prediction credibility of the denoised left neighborhood electric energy data point sequence and the prediction credibility of the denoised right neighborhood electric energy data point sequence based on the analysis results, and determine the prediction weight value based on the obtained prediction credibility. The prediction weight value can reflect the contribution of the prediction results in different directions to the final prediction result, that is, the prediction result with high prediction credibility has a greater contribution, thereby improving the prediction accuracy. Based on the above analysis, it can be seen that this embodiment will first The prediction credibility of the denoised left neighboring electric energy data point sequence is obtained by performing time series decomposition on the trend component data point sequence and the periodic component data point sequence, and the peak point on the frequency domain signal obtained by performing frequency domain conversion on the periodic component data point sequence, and the prediction credibility of the denoised right neighboring electric energy data point sequence is obtained according to the trend component data point sequence and the periodic component data point sequence obtained by performing time series decomposition on the denoised right neighboring electric energy data point sequence, and the peak point on the frequency domain signal obtained by performing frequency domain conversion on the periodic component data point sequence; and since the method for obtaining the prediction credibility of the denoised right neighboring electric energy data point sequence is the same as the method for obtaining the prediction credibility of the denoised left neighboring electric energy data point sequence, this embodiment will only describe the process of obtaining the prediction credibility of the denoised left neighboring electric energy data point sequence in the following. Then the specific process of obtaining the prediction credibility of the denoised left neighboring electric energy data point sequence is as follows: First, the STL time series decomposition is performed on the denoised left neighborhood electric energy data point sequence to obtain the trend component data point sequence and the period component data point sequence corresponding to the denoised left neighborhood electric energy data point sequence, and the trend component data point sequence corresponding to the denoised left neighborhood electric energy data point sequence is recorded as the first trend sequence, and the period component data point sequence corresponding to the denoised left neighborhood electric energy data point sequence is recorded as the first period sequence. The STL time series decomposition of the denoised left neighborhood electric energy data point sequence is actually the STL time series decomposition of the time series sequence composed of the ordinates of all electric energy data points in the denoised left neighborhood electric energy data point sequence. The ordinate is the electric energy data, and the abscissa value of the hth trend component data point in the trend component data point sequence corresponding to the denoised left neighborhood electric energy data point sequence is the denoised left neighborhood electric energy data point sequence. The horizontal coordinate value of the h-th electric energy data point in the data point sequence and the vertical coordinate value of the h-th trend component data point are the trend components obtained by performing STL time series decomposition on the vertical coordinate value of the h-th electric energy data point in the denoised left neighboring electric energy data point sequence. The horizontal coordinate value of the h-th periodic component data point in the periodic component data point sequence corresponding to the denoised left neighboring electric energy data point sequence is the horizontal coordinate value of the h-th electric energy data point in the denoised left neighboring electric energy data point sequence, and the vertical coordinate value of the h-th periodic component data point is the periodic component obtained by performing STL time series decomposition on the vertical coordinate value of the h-th electric energy data point in the denoised left neighboring electric energy data point sequence. The horizontal coordinates of all points in this embodiment represent time. Since the process of performing time series decomposition is well known, this embodiment will not be described in detail.

[0022] After obtaining the first trend sequence, the inflection point in the first trend sequence is obtained, and the process of obtaining the inflection point is well known; then, based on the distance between the trend component data point in the first trend sequence and the adjacent trend component data point and the inflection point in the first trend sequence, the trend item stable eigenvalue of the denoised left neighborhood electric energy data point sequence is obtained. The trend item stable eigenvalue is a key indicator for subsequently determining the prediction credibility of the denoised left neighborhood electric energy data point sequence. Then, the specific process of obtaining the trend item stable eigenvalue of the denoised left neighborhood electric energy data point sequence is as follows: first, based on the Euclidean distance between each trend component data point in the first trend sequence and the adjacent trend component data point of the corresponding trend component data point, the neighborhood trend distance of each trend component data point in the first trend sequence is obtained; then, the neighborhood slope difference value of each inflection point in the first trend sequence is obtained, and based on the neighborhood trend distance of the trend component data point in the first trend sequence and the neighborhood slope difference value of the inflection point, the trend item stable eigenvalue of the denoised left neighborhood electric energy data point sequence is obtained.

[0023] In this embodiment, the specific process of obtaining the neighborhood trend distance of the trend component data point is as follows: for the kth trend component data point in the first trend sequence, if k is 1, the Euclidean distance between the kth trend component data point and the k+1th trend component data point in the first trend sequence is used as the neighborhood trend distance of the kth trend component data point; if k is K0, the Euclidean distance between the k-1th trend component data point and the kth trend component data point in the first trend sequence is used as the neighborhood trend distance of the kth trend component data point; if k is not equal to 1 and K0, the Euclidean distance between the kth trend component data point and the k-1th trend component data point in the first trend sequence is used as the neighborhood trend distance of the kth trend component data point. The Euclidean distance between potential component data points plus the Euclidean distance between the kth trend component data point and the k+1th trend component data point in the first trend sequence is taken as the neighborhood trend distance of the kth trend component data point, that is, if k is not equal to 1 and K0, then the sum of D3 and D4 is the neighborhood electric energy distance of the kth trend component data point, K0 is the total number of trend component data points in the first trend sequence, D3 is the Euclidean distance between the kth trend component data point and the k-1th trend component data point in the first trend sequence, and D4 is the Euclidean distance between the kth trend component data point and the k+1th trend component data point in the first trend sequence.

[0024] In this embodiment, the specific process of obtaining the neighborhood slope difference value of each inflection point in the first trend sequence is: for any inflection point in the first trend sequence, in the first trend sequence, obtain the trend component data point located on the left side of the inflection point and adjacent to the inflection point, and record it as the left adjacent trend component data point of the inflection point, obtain the trend component data point located on the right side of the inflection point and adjacent to the inflection point, and record it as the right adjacent trend component data point of the inflection point, calculate the slope between the left adjacent trend component data point of the inflection point and the inflection point, and record it as the first slope of the inflection point, calculate the slope between the right adjacent trend component data point of the inflection point and the inflection point, and record it as the second slope of the inflection point, calculate the absolute value of the difference between the first slope and the second slope of the inflection point, and use it as the neighborhood slope difference value of the inflection point; the calculation method of the slope between any two points is a well-known technology.

[0025] In this embodiment, the specific process of obtaining the stable characteristic value of the trend item of the denoised left neighborhood electric energy data point sequence according to the neighborhood trend distance of the trend component data point and the neighborhood slope difference value of the inflection point in the first trend sequence is as follows: calculating the mean of the neighborhood trend distances of all trend component data points in the first trend sequence and recording it as the neighborhood trend distance mean, calculating the cumulative result of the neighborhood slope difference values ​​of all inflection points in the first trend sequence and recording it as the comprehensive neighborhood slope difference value, multiplying the neighborhood trend distance mean by the comprehensive neighborhood slope difference value, and performing negative correlation mapping on the multiplication result, and recording the mapping result as the trend item stable characteristic value of the denoised left neighborhood electric energy data point sequence; and the calculation expression of the stable characteristic value of the trend item of the denoised left neighborhood electric energy data point sequence is:

[0026] Among them, R1 is the stable eigenvalue of the trend term of the denoised left neighborhood electric energy data point sequence, is the neighborhood trend distance of the kth trend component data point in the first trend sequence, K0 is the total number of trend component data points in the first trend sequence, K1 is the number of inflection points in the first trend sequence, is the first slope of the i-th inflection point in the first trend sequence, is the second slope of the i-th inflection point in the first trend sequence, exp() is an exponential function with a constant e as the base, and the role of exp() here is to perform negative correlation mapping. And when The smaller the value is, the more stable the trend term of the denoised left neighborhood electric energy data point sequence is. The smaller it is, the smoother the curve formed by the data points in the first trend sequence of the denoised left neighborhood electric energy data point sequence is. When the overall trend item of the denoised left neighborhood electric energy data point sequence is more stable and the curve formed by the data points in the first trend sequence of the denoised left neighborhood electric energy data point sequence is smoother, it indicates that the continuous similarity feature between the distribution trend of the electric energy data in the left neighborhood time stage and the distribution trend of the electric energy data in the power-off time stage is more obvious. When the continuous similarity feature is more obvious, the regularity stability of the electric energy data in the left neighborhood time stage is higher, that is, the regularity stability of the denoised left neighborhood electric energy data point sequence is higher. When the regularity stability of the denoised left neighborhood electric energy data point sequence is higher, the credibility of the prediction result based on the denoised left neighborhood electric energy data point sequence is higher. Therefore, when The smaller it is, that is, the larger R1 is, the higher the credibility of the prediction result based on the denoised left neighborhood electric energy data point sequence is. Conversely, the smaller R1 is, the lower the credibility of the prediction result based on the denoised left neighborhood electric energy data point sequence is.

[0027] After obtaining the stable characteristic value of the trend item, the first periodic sequence is subjected to Fourier transform, and the frequency domain signal obtained by the Fourier transform is recorded as the first frequency domain signal. Then, the frequency corresponding to the peak point with the largest amplitude on the first frequency domain signal is obtained and recorded as the main frequency. The inverse of the main frequency is obtained and the inverse of the main frequency is taken as the first main period. Then, the first periodic sequence is divided by the first main period to obtain various subsequences on the first periodic sequence. The purpose of dividing the first periodic sequence is to analyze the stability of the periodic term, and the stability of the periodic term is a key indicator for the subsequent determination of the prediction credibility of the denoised left neighborhood electric energy data point sequence. In addition, an example of using the first main period to divide the first periodic sequence to obtain various subsequences on the first periodic sequence is as follows: If the time length of the first main period is T, then the periodic component data points in the first periodic sequence are first connected in chronological order, and then The connected curve is recorded as the first periodic curve. Starting from the starting periodic component data point on the first periodic curve, the first periodic curve is divided into non-overlapping segments using the time length T to obtain all sub-curve segments on the first periodic curve. The sequence consisting of all periodic component data points on each sub-curve segment in chronological order is recorded as the subsequence corresponding to the corresponding sub-curve segment, also recorded as each subsequence on the first periodic sequence. That is, the subsequence corresponding to the j-th sub-curve segment on the first periodic curve is the j-th subsequence on the first periodic sequence. In addition, except for the last subsequence on the first periodic sequence, the time length corresponding to other subsequences is T, and the time length corresponding to the last divided time period on the first periodic sequence is less than or equal to T. The time length corresponding to a subsequence refers to the time interval between the starting point and the last point on the sub-curve segment corresponding to the subsequence, which is also the horizontal coordinate difference.

[0028] Then, based on the periodic component differences between the periodic component data points at the same position in adjacent subsequences on the first periodic sequence, the periodic term stable eigenvalue of the denoised left neighborhood electric energy data point sequence is obtained; and the specific calculation process for obtaining the periodic term stable eigenvalue of the denoised left neighborhood electric energy data point sequence is as follows: First, the neighborhood periodic component difference value of each periodic component data point in each subsequence on the first periodic sequence is obtained. Then, for any subsequence, the cumulative sum of the neighborhood periodic component difference values ​​corresponding to all periodic component data points in the subsequence is recorded as the comprehensive neighborhood periodic component difference value of the subsequence. The mean of the comprehensive neighborhood periodic component difference values ​​of all subsequences on the first periodic sequence is calculated and recorded as the comprehensive mean. The comprehensive mean is negatively correlated and mapped, and the mapping result is recorded as the stable eigenvalue of the periodic term corresponding to the denoised left neighborhood electric energy data point sequence.

[0029] In this embodiment, the specific process of obtaining the neighborhood periodic component difference value of each periodic component data point in each subsequence on the first periodic sequence is as follows: for the fth periodic component data point in the dth subsequence on the first periodic sequence: if d is 1, then the absolute value of the vertical coordinate difference between the fth periodic component data point in the dth subsequence and the fth periodic component data point in the d+1th subsequence on the first periodic sequence is used as the neighborhood periodic component difference value of the fth periodic component data point in the dth subsequence; if d is D, then the fth periodic component data point in the dth subsequence is used as the neighborhood periodic component difference value of the fth periodic component data point. The absolute value of the difference in ordinate between the data point and the f-th period component data point in the d-1th subsequence of the first periodic sequence is used as the neighborhood periodic component difference value of the f-th period component data point in the d-th subsequence; if d is not 1 or D, the absolute value of the difference in ordinate between the f-th period component data point in the d-th subsequence and the f-th period component data point in the d+1th subsequence is added to the absolute value of the difference in ordinate between the f-th period component data point in the d-1th subsequence and the f-th period component data point in the d-1th subsequence, and the result is used as the neighborhood periodic component difference value of the f-th period component data point in the d-th subsequence. The neighborhood periodic component difference value of the f-th periodic component data point, that is, if d is not 1 and D, then the neighborhood periodic component difference value of the f-th periodic component data point in the d-th subsequence is the sum of F1 and F2, F1 is the absolute value of the vertical coordinate difference between the f-th periodic component data point in the d-th subsequence and the f-th periodic component data point on the d+1-th subsequence, F2 is the absolute value of the vertical coordinate difference between the f-th periodic component data point in the d-th subsequence and the f-th periodic component data point in the d-1-th subsequence, the horizontal coordinate of the periodic component data point is time, and the vertical coordinate is period division. where D is the number of subsequences on the first periodic sequence; and if the time length corresponding to the last subsequence on the first periodic sequence is less than the first main period, the d-th subsequence is the second-to-last subsequence on the first periodic sequence, and the number of periodic component data points in the last subsequence on the first periodic sequence is less than f, then the neighborhood periodic component difference value of the f-th periodic component data point on the d-th subsequence is also the absolute value of the vertical coordinate difference between the f-th periodic component data point in the d-th subsequence and the f-th periodic component data point in the d-1-th subsequence on the first periodic sequence.

[0030] In addition, the specific calculation expression of the stable eigenvalue of the periodic term corresponding to the denoised left neighborhood electric energy data point sequence is:

[0031] Where R2 is the stable eigenvalue of the periodic term corresponding to the denoised left neighborhood electric energy data point sequence, D is the number of subsequences on the first periodic sequence, is the comprehensive neighborhood periodic component difference value of the d-th subsequence on the first periodic sequence, and the role of exp() here is to perform negative correlation mapping; because when the adjacent subsequences, that is, the adjacent sub-curve segments are more similar, the regularity stability of the denoised left neighborhood electric energy data point sequence is higher, and when the regularity stability of the denoised left neighborhood electric energy data point sequence is higher, the credibility of the prediction result based on the denoised left neighborhood electric energy data point sequence is higher, and because when The smaller the value, the higher the similarity between adjacent subsequences on the first periodic sequence. The smaller the value, the larger the R2. Therefore, in this embodiment, when R2 is larger, it indicates that the similarity between adjacent subsequences on the first periodic sequence is higher, indicating that the regularity and stability of the denoised left neighborhood electric energy data point sequence is higher, and also indicating that the credibility of the prediction result obtained based on the denoised left neighborhood electric energy data point sequence is higher. Conversely, when R2 is smaller, it indicates that the credibility of the prediction result obtained based on the denoised left neighborhood electric energy data point sequence is lower.

[0032] After obtaining the stable eigenvalues ​​of the periodic term and the stable eigenvalues ​​of the trend term of the denoised left neighborhood electric energy data point sequence, the sum of the stable eigenvalues ​​of the periodic term and the stable eigenvalues ​​of the trend term of the denoised left neighborhood electric energy data point sequence is used as the prediction credibility of the denoised left neighborhood electric energy data point sequence. That is, the larger the stable eigenvalues ​​of the periodic term and the stable eigenvalues ​​of the trend term or the larger the prediction credibility of the denoised left neighborhood electric energy data point sequence, the higher the reliability or credibility of the predicted electric energy in the power outage time stage obtained based on the denoised left neighborhood electric energy data point sequence; conversely, the lower the reliability or credibility of the predicted electric energy in the power outage time stage obtained based on the denoised left neighborhood electric energy data point sequence.

[0033] After obtaining the prediction credibility of the denoised left neighborhood electric energy data point sequence and the prediction credibility of the denoised right neighborhood electric energy data point sequence, the prediction weight value is obtained by summing and normalizing. The greater the prediction credibility of the denoised left neighborhood electric energy data point sequence, the greater the prediction weight value of the denoised left neighborhood electric energy data point sequence obtained by normalization. The greater the prediction weight value of the denoised neighborhood electric energy data point sequence, the higher the contribution or participation of the predicted electric energy in the power outage time stage obtained based on the denoised neighborhood electric energy data point sequence in the final determination of the predicted electric energy in the power outage time stage. Then, according to the prediction credibility of the denoised left neighborhood electric energy data point sequence and the denoised right neighborhood electric energy data point sequence, the specific process of obtaining the prediction weight value of the denoised left neighborhood electric energy data point sequence and the prediction weight value of the denoised right neighborhood electric energy data point sequence is as follows: Calculate the sum of the prediction credibility of the denoised left neighborhood electric energy data point sequence and the prediction credibility of the denoised right neighborhood electric energy data point sequence, and record it as the comprehensive prediction credibility. Calculate the ratio of the prediction credibility of the denoised left neighborhood electric energy data point sequence to the comprehensive prediction credibility, and record it as the prediction weight value of the denoised left neighborhood electric energy data point sequence. Calculate the ratio of the prediction credibility of the denoised right neighborhood electric energy data point sequence to the comprehensive prediction credibility, and record it as the prediction weight value of the denoised right neighborhood electric energy data point sequence.

[0034] Therefore, this embodiment can obtain the prediction weight value of the denoised left neighborhood electric energy data point sequence and the prediction weight value of the denoised right neighborhood electric energy data point sequence through the above process.

[0035] The prediction module 03 is configured to obtain target predicted electric energy data in the power outage period according to the prediction weight values ​​of the denoised left neighborhood electric energy data point sequence and the denoised right neighborhood electric energy data point sequence.

[0036] After obtaining the prediction weight value of the denoised left neighborhood electric energy data point sequence and the prediction weight value of the denoised right neighborhood electric energy data point sequence, the electric energy during the power outage time period is predicted based on the denoised left neighborhood electric energy data point sequence using the ARIMA prediction algorithm, and the obtained prediction data are all recorded as the first predicted electric energy data. According to the denoised right neighborhood electric energy data point sequence, the electric energy during the power outage time period is predicted using the ARIMA prediction algorithm, and the obtained prediction data are all recorded as the second predicted electric energy data, that is, the first predicted electric energy data is predicted based on the denoised left neighborhood electric energy data point sequence, and the second predicted electric energy data is predicted based on the denoised right neighborhood electric energy data point sequence; since the process of using the ARIMA prediction algorithm for forward or reverse data prediction is a well-known technology under the premise of knowing the data used in the prediction, it will not be described in detail in this embodiment.

[0037] Then, according to the predicted weight value of the denoised left neighborhood electric energy data point sequence and the predicted weight value of the denoised right neighborhood electric energy data point sequence, the obtained first predicted electric energy data and the second predicted electric energy data are weightedly fused to obtain the target predicted electric energy data in the power outage time stage. The specific process is: for any power outage moment in the power outage time stage: the result of multiplying the first predicted electric energy data at the power outage moment by the predicted weight value of the denoised left neighborhood electric energy data point sequence is recorded as the first weighted data, the result of multiplying the second predicted electric energy data at the power outage moment by the predicted weight value of the denoised right neighborhood electric energy data point sequence is recorded as the second weighted data, and the sum of the first weighted data and the second weighted data is recorded as the target predicted electric energy data at the power outage moment.

[0038] Therefore, this embodiment can obtain the target predicted electric energy data at each power outage moment in the power outage time stage through the above method, and finally supplement and insert the target predicted electric energy data at each power outage moment into the corresponding power outage moment position, thereby obtaining the electric energy data of the power outage time stage.

[0039] At this point, this embodiment completes the prediction and interpolation of the electric energy data during the power outage period.

[0040] To summarize, this embodiment includes a data acquisition module for acquiring a denoised neighborhood electric energy data point sequence corresponding to the power outage time stage of the concentrator; a prediction weight acquisition module for obtaining the prediction credibility of the denoised neighborhood electric energy data point sequence based on the trend component data point sequence and the periodic component data point sequence obtained by time series decomposition of the denoised neighborhood electric energy data point sequence and the peak points on the frequency domain signal obtained by frequency domain conversion of the periodic component data point sequence, and obtaining the prediction weight value of the denoised left neighborhood electric energy data point sequence and the prediction weight value of the denoised right neighborhood electric energy data point sequence based on the prediction credibility of the denoised left neighborhood electric energy data point sequence and the denoised right neighborhood electric energy data point sequence; a prediction module for obtaining the target predicted electric energy data in the power outage time stage based on the prediction weight values ​​of the denoised left neighborhood electric energy data point sequence and the denoised right neighborhood electric energy data point sequence. Moreover, this embodiment is based on the prediction weight values ​​of the denoised left neighboring electric energy data point sequence and the denoised right neighboring electric energy data point sequence and the denoised left neighboring electric energy data point sequence and the denoised right neighboring electric energy data point sequence, which can improve the accuracy of electric energy prediction during the power outage time period, thereby improving the accuracy of electric energy prediction interpolation during the power outage time period.

[0041] The above-described embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present application, and should all be included in the scope of protection of the present application.

Claims

1. A concentrator with intelligent power outage recovery function, characterized in that: The concentrator with power failure intelligent recovery function includes: A data acquisition module is used to acquire a denoised neighborhood electric energy data point sequence corresponding to the power outage time stage of the concentrator, wherein the denoised neighborhood electric energy data point sequence includes a denoised left neighborhood electric energy data point sequence and a denoised right neighborhood electric energy data point sequence; a prediction weight acquisition module, configured to obtain the prediction credibility of the denoised neighborhood electric energy data point sequence based on the trend component data point sequence and the periodic component data point sequence obtained by time series decomposition of the denoised neighborhood electric energy data point sequence, and the peak points on the frequency domain signal obtained by frequency domain conversion of the periodic component data point sequence, and to obtain the prediction weight value of the denoised left neighborhood electric energy data point sequence and the prediction weight value of the denoised right neighborhood electric energy data point sequence based on the prediction credibility of the denoised left neighborhood electric energy data point sequence and the denoised right neighborhood electric energy data point sequence; The prediction module is used to obtain the target predicted electric energy data in the power outage time stage according to the prediction weight values ​​of the denoised left neighborhood electric energy data point sequence and the denoised right neighborhood electric energy data point sequence.

2. The concentrator with power failure intelligent recovery function according to claim 1, characterized in that: The method for obtaining the denoised neighborhood electric energy data point sequence corresponding to the power outage time stage includes: Obtaining a left neighboring electric energy data point sequence corresponding to a left neighboring time stage and a right neighboring electric energy data point sequence corresponding to a right neighboring time stage of the power outage time stage; According to the distance between the electric energy data point in the left neighborhood electric energy data point sequence and the adjacent electric energy data point of the corresponding electric energy data point, the noise data point in the left neighborhood electric energy data point sequence is obtained, and the noise data point in the left neighborhood electric energy data point sequence is interpolated and replaced to obtain a denoised left neighborhood electric energy data point sequence; the method for obtaining the denoised right neighborhood electric energy data point sequence is the same as the method for obtaining the denoised left neighborhood electric energy data point sequence, and the denoised right neighborhood electric energy data point sequence and the denoised left neighborhood electric energy data point sequence both belong to the denoised neighborhood electric energy data point sequence.

3. The concentrator with power failure intelligent recovery function as claimed in claim 2, characterized in that: The method for obtaining noise data points in the left-neighborhood electric energy data point sequence comprises: Obtaining a neighborhood electric energy distance of each electric energy data point in the left neighborhood electric energy data point sequence, where the neighborhood electric energy distance of the electric energy data point is the Euclidean distance between the electric energy data point and an adjacent electric energy data point of the corresponding electric energy data point; For the ath electric energy data point in the left neighborhood electric energy data point sequence, a is greater than 1 and less than A, and A is the total number of electric energy data points in the left neighborhood electric energy data point sequence: the absolute value of the difference between the neighborhood electric energy distance of the a-1th electric energy data point in the left neighborhood electric energy data point sequence and the neighborhood electric energy distance of the ath electric energy data point is added to the absolute value of the difference between the neighborhood electric energy distance of the a+1th electric energy data point in the left neighborhood electric energy data point sequence and the neighborhood electric energy distance of the ath electric energy data point, and then the result is normalized and recorded as a comprehensive difference; the sum of the normalized value of the neighborhood electric energy distance of the ath electric energy data point and the comprehensive difference is used as the noise index value of the ath electric energy data point; if the noise index value is greater than the preset noise threshold, the ath electric energy data point is recorded as a noise data point.

4. The concentrator with power failure intelligent recovery function according to claim 1, characterized in that: Methods for obtaining prediction credibility include: Recording a trend component data point sequence and a period component data point sequence obtained by performing time series decomposition on the denoised left-neighborhood electric energy data point sequence as a first trend sequence and a first period sequence, respectively; recording a frequency domain signal obtained by performing Fourier transform on the first period sequence as a first frequency domain signal; recording the frequency of a peak point with the largest amplitude on the first frequency domain signal as a main frequency; taking the reciprocal of the main frequency as a first main period; and dividing the first period sequence using the first main period to obtain subsequences on the first period sequence; Obtaining a stable eigenvalue of a trend term of the denoised left-neighborhood electric energy data point sequence based on a distance between a trend component data point and an adjacent trend component data point in the first trend sequence and an inflection point in the first trend sequence, obtaining a stable eigenvalue of a period term of the denoised left-neighborhood electric energy data point sequence based on a difference in periodic components between periodic component data points at the same position in adjacent subsequences of the first periodic sequence, and using the sum of the stable eigenvalue of the trend term and the stable eigenvalue of the period term as the prediction credibility of the denoised left-neighborhood electric energy data point sequence; The method for obtaining the prediction credibility of the denoised right neighborhood electric energy data point sequence is the same as the method for obtaining the prediction credibility of the denoised left neighborhood electric energy data point sequence.

5. The concentrator with power failure intelligent recovery function as claimed in claim 4, characterized in that: The method for obtaining the stable characteristic value of the trend item of the denoised left neighborhood electric energy data point sequence comprises: The Euclidean distance between each trend component data point in the first trend sequence and the adjacent trend component data point of the corresponding trend component data point is used as the neighborhood trend distance of the corresponding trend component data point; the neighborhood slope difference value of each inflection point in the first trend sequence is obtained, and the neighborhood slope difference value of any inflection point in the first trend sequence is the absolute value of the difference between the first slope and the second slope of the inflection point, where the first slope of the inflection point is the slope between the adjacent trend component data point on the left side of the inflection point and the inflection point, and the second slope of the inflection point is the slope between the adjacent trend component data point on the right side of the inflection point and the inflection point; According to the neighborhood trend distance of the trend component data point and the neighborhood slope difference value of the inflection point in the first trend sequence, the trend term stable eigenvalue of the denoised left neighborhood electric energy data point sequence is obtained.

6. The concentrator with intelligent power-off recovery function according to claim 5, characterized in that: The method for obtaining a stable characteristic value of a trend item of the denoised left neighborhood electric energy data point sequence according to a neighborhood trend distance of a trend component data point and a neighborhood slope difference value of an inflection point in a first trend sequence comprises: The mean of the neighborhood trend distances of all trend component data points in the first trend sequence is recorded as the neighborhood trend distance mean, the cumulative result of the neighborhood slope difference values ​​of all inflection points in the first trend sequence is recorded as the comprehensive neighborhood slope difference value, and the result of multiplying the neighborhood trend distance mean by the comprehensive neighborhood slope difference value and then performing negative correlation mapping is recorded as the trend item stable eigenvalue of the denoised left neighborhood electric energy data point sequence.

7. The concentrator with power failure intelligent recovery function according to claim 4, characterized in that: The method for obtaining the stable eigenvalue of the periodic term of the denoised left-neighborhood electric energy data point sequence comprises: For the f-th periodic component data point in the d-th subsequence of the first periodic sequence, obtain a neighborhood periodic component difference value of the f-th periodic component data point in the d-th subsequence based on the difference between the f-th periodic component data point in the subsequence adjacent to the d-th subsequence and the f-th periodic component data point in the d-th subsequence; The cumulative sum of the neighborhood periodic component difference values ​​of all periodic component data points in each subsequence is recorded as the comprehensive neighborhood periodic component difference value of the corresponding subsequence, and the result of negative correlation mapping of the mean of the comprehensive neighborhood periodic component difference values ​​of all subsequences on the first periodic sequence is recorded as the stable eigenvalue of the periodic term corresponding to the denoised left neighborhood electric energy data point sequence.

8. The concentrator with power failure intelligent recovery function according to claim 7, characterized in that: The method for obtaining the neighborhood periodic component difference value of the f-th periodic component data point in the d-th subsequence includes: If d is 1, the absolute value of the difference in ordinate between the f-th periodic component data point in the d-th subsequence and the f-th periodic component data point in the d+1-th subsequence on the first periodic sequence is used as the neighborhood periodic component difference value of the f-th periodic component data point in the d-th subsequence; if d is D, the absolute value of the difference in ordinate between the f-th periodic component data point in the d-th subsequence and the f-th periodic component data point in the d-1-th subsequence on the first periodic sequence is used as the neighborhood periodic component difference value of the f-th periodic component data point on the d-th subsequence. If d is not 1 and D, the absolute value of the vertical coordinate difference between the f-th periodic component data point in the d-th subsequence and the f-th periodic component data point in the d+1-th subsequence is added to the absolute value of the vertical coordinate difference between the f-th periodic component data point in the d-th subsequence and the f-th periodic component data point in the d-1-th subsequence, as the neighborhood periodic component difference value of the f-th periodic component data point in the d-th subsequence, the horizontal coordinate of the periodic component data point is time, the vertical coordinate is the periodic component, and D is the number of subsequences in the first periodic sequence.

9. The concentrator with intelligent power-off recovery function according to claim 1, characterized in that: The prediction weight value of the denoised left neighborhood electric energy data point sequence and the prediction weight value of the denoised right neighborhood electric energy data point sequence are the results of summing and normalizing the prediction credibility of the denoised left neighborhood electric energy data point sequence and the prediction credibility of the denoised right neighborhood electric energy data point sequence.

10. The concentrator with power failure intelligent recovery function according to claim 1, characterized in that: The method for obtaining target predicted electric energy data includes: The electric energy during the power outage period is predicted based on the denoised left-neighborhood electric energy data point sequence to obtain first predicted electric energy data for the power outage period; the electric energy during the power outage period is predicted based on the denoised right-neighborhood electric energy data point sequence to obtain second predicted electric energy data for the power outage period; For any power outage moment in the power outage time period, the result of multiplying the first predicted electric energy data at the power outage moment by the predicted weight value of the denoised left neighboring electric energy data point sequence is recorded as the first weighted data, the result of multiplying the second predicted electric energy data at the power outage moment by the predicted weight value of the denoised right neighboring electric energy data point sequence is recorded as the second weighted data, and the sum of the first weighted data and the second weighted data is recorded as the target predicted electric energy data at the power outage moment.

Images