Load control strategy generation method and system suitable for regional power grid

Abstract

The invention discloses a load control strategy generation method and system suitable for a regional power grid. The method comprises the following steps: acquiring power utilization historical active power data of a load user, and preprocessing the active power data to form a preprocessed power utilization data set; carrying out segmentation processing on the preprocessed power consumption data set of the user according to a time stage and a specific user to generate a time sequence, and carrying out dynamic adjustment according to the time stage and behavior characteristics of the specific user; and converting the time sequence into a frequency domain space, and identifying peak valleys and fluctuation modes in a power consumption curve of the user. Meanwhile, geographic features, industry features, scale features and load characteristics are considered, and load control strategies meeting various user requirements are generated by combining user classification results obtained through clustering analysis and other quantized key features. According to the strategy generation method, pertinence and adaptability of load control are improved, and a power supply company can be helped to better adjust power supply and demand contradictory demands.

Description

Technical Field

[0001] The present invention relates to the technical field of power load management, and in particular to a method and system for generating a load control strategy suitable for a regional power grid. Background Art

[0002] With the continuous growth of energy demand and the expansion of power grid scale, in order to alleviate the seasonal and periodic contradiction between power supply and demand, power supply companies adopt load control measures such as peak shifting, peak avoidance, rotation, power supply and load power restriction to reduce load. Traditional load control strategies have certain limitations in meeting the characteristics of different regions and industry characteristics. Traditional methods usually adopt load distribution based on manual experience, ignoring the differences in power consumption, economic structure and environmental conditions in different regions. Therefore, a load control strategy generation method suitable for regional power grids is needed to reasonably specify load control strategy solutions. At present, although there are some regionalized load management methods, they are often compiled and generated based on manual experience, and lack statistical analysis and judgment of data. Summary of the invention

[0003] Purpose of the invention: In order to overcome the shortcomings of the prior art, the present invention provides a method for generating a load control strategy suitable for a regional power grid. The method solves the problem of insufficient stability of the power supply system caused by poor accuracy of the load control strategy. The present invention also provides a system for generating a load control strategy suitable for a regional power grid.

[0004] Technical solution: In a first aspect, the present invention provides a method for generating a load control strategy applicable to a regional power grid, the method comprising:

[0005] Obtaining historical active power data of load users, and preprocessing the active power data to form a preprocessed power consumption data set, wherein the active power data includes data collected by a smart meter or a remote collection system;

[0006] The pre-processed user power consumption data set is segmented according to time stages and specific users to generate a time series, and dynamically adjusted according to the time stages and the behavioral characteristics of the specific users;

[0007] Convert the time series into frequency domain space and identify peaks, valleys and fluctuation patterns in the user's power consumption curve;

[0008] Determine the type of power load by the peaks, valleys and fluctuation patterns in the power consumption curve, and adaptively cluster the power load type, and the characteristic vector of the adaptive clustering includes: peak information during the morning peak period and evening peak period of each day; valley information corresponding to all time points of each day; and the average value of the corresponding information of the daily user curve, thereby generating a suitable clustering area;

[0009] The user classification results obtained by cluster analysis are combined with other quantified key features to generate load control strategies that meet the needs of various types of users. The other key features include geographical characteristics, industry characteristics, scale characteristics, load characteristics and load elasticity characteristics.

[0010] Further, including:

[0011] The acquisition of the load user's historical active power data includes: continuously recording and storing the user's daily, hourly, and minute power consumption, and high-frequency continuous collection and recording of power.

[0012] Further, including:

[0013] The preprocessing of the active power data includes determining whether the active power data of the user at the moment is abnormal according to the power change rate between adjacent or two fixed time points, and if it is abnormal data, deleting the relevant data.

[0014] Further, including:

[0015] The determining whether the active power data of the user at the moment is abnormal according to the power change rate between adjacent or two fixed time points includes:

[0016] Traverse the collected historical active power data and corresponding time, and use the active power data at the same time in a certain period of time as a new data set, that is, form multiple data sets formed at fixed time points every day;

[0017] Taking each data set formed at the same fixed time point as the benchmark, the difference of active power data of adjacent data is calculated. If the difference between different days at the same time is not greater than the first threshold, the data are deemed valid; otherwise,

[0018] If the current difference is greater than the first threshold, the time corresponding to the two adjacent data is checked. If the data change is caused by climate, mainly temperature, the current two data are retained. Otherwise, the current data of the next day is deemed invalid, and the difference between the data of the next two days and the current data is calculated and iterative judgment is performed.

[0019] Taking each data set formed at different time points on the same day as a benchmark, the difference of active power data of adjacent data is calculated. If the difference calculated at different times on the same day is not greater than the second threshold, the data are deemed valid; otherwise,

[0020] If the current difference is greater than the second threshold, the time corresponding to the two adjacent data is checked. If the data change is caused by day and night, the current two data are retained. Otherwise, the current data of the next day is deemed to be an invalid value, and the difference between the data of the next two days and the current data is continued to be calculated and then iterative judgment is made.

[0021] Further, including:

[0022] The identifying of peaks, valleys and fluctuation patterns in the user's electricity consumption curve includes:

[0023] Dynamically identify non-periodic fluctuation patterns through adaptive algorithms;

[0024] According to the peak and valley conditions of the user's electricity consumption curve, the peak information corresponding to the electricity consumption curve is extracted to determine three types of electricity load, including: daytime double hump, reverse peak type and continuous type. The daytime double hump refers to the user's electricity consumption behavior being manifested in that the load electricity consumption curve has an obvious hump feature, that is, there are two electricity consumption peaks; the reverse peak type refers to the user's electricity consumption behavior being manifested in that there is no obvious load during the day, but there is a peak of the production work plan at night; the continuous type refers to the user's electricity consumption behavior being relatively continuous, without an obvious peak, that is, the corresponding load curve fluctuates slightly within a day, and has no obvious peaks or troughs, and it is determined to be a continuous type.

[0025] Further, including:

[0026] The method of determining the type of power load by using the peaks, valleys and fluctuation patterns in the power consumption curve and adaptively clustering the power load type comprises:

[0027] The relevant load characteristics of each user are extracted as a vector. The relevant load characteristics include: peak information during the morning peak period and the evening peak period; valley information corresponding to all time points of the day; and the average value of the corresponding information of the daily user curve; three clustering centers corresponding to daytime double hump, anti-peak type and continuous type are selected, and the clustering algorithm is used to classify the feature vectors; users are preliminarily divided into three categories: daytime double hump, anti-peak type and continuous type.

[0028] Further, including:

[0029] The other key characteristics include: geographical characteristics, industry characteristics, scale characteristics, load characteristics and load elasticity characteristics, and the other key characteristics are quantified;

[0030] The geographic feature G of the load user is processed as follows: the latitude and longitude position coordinates of the load user are quantized to form a geographic feature matrix, and a first constraint condition is set;

[0031] The industry characteristics I of the load users are processed as follows: the ratio of electricity cost to output value of each industry is calculated to form an industry characteristics matrix, and a second constraint condition on economic benefits is introduced;

[0032] The scale characteristic L of the load user is processed as follows: considering the requirements of the fast up and fast down group and the load control group of the load control strategy, setting the third constraint condition to put the bimodal users into the fast up and fast down group;

[0033] The scale characteristics S of the load users are processed as follows: the users are quantified according to their electricity consumption scale, and the users are divided into two main categories: large-scale enterprises and small-scale enterprises according to their electricity consumption and electricity demand. The third constraint condition is introduced, and different weight parameters are set to select load users of different scales;

[0034] The load elasticity characteristic E of the load user is processed as follows: the load adjustment elasticity of the user is calculated according to the historical response behavior of the user to form an elasticity characteristic matrix.

[0035] Further, including:

[0036] Combining the user classification results obtained by cluster analysis with other quantified key features, a load control strategy that meets the needs of various users is generated, including:

[0037] A corresponding feature vector is formed according to the geographical characteristics, industry characteristics, scale characteristics, load characteristics and load elasticity characteristics, and a priority weight is defined for the feature vector of each user; in each time period, the power usage right is allocated to the user with the largest selection weight, and the load of the power grid and the weight of the user are updated; this process is repeated until all users are allocated power usage rights, and the final load control plan is determined by comprehensively considering the results of the user power usage rights allocation.

[0038] Further, including:

[0039] The active power data of each user is divided into stages according to a certain period of time. The load type, characteristic vector and priority weight corresponding to each user are taken as input, and the greedy algorithm is used to allocate electricity rights to each user.

[0040] In a second aspect, the present invention further provides a load control strategy generation system applicable to a regional power grid, the system comprising:

[0041] A data preprocessing module is used to obtain the historical active power data of the load user and preprocess the active power data to form a preprocessed power data set, wherein the active power data includes data collected by a smart meter or a remote collection system;

[0042] The power consumption curve recognition module is used to segment the pre-processed user power consumption data set according to the time stage and specific user, generate a time series, and dynamically adjust it according to the time stage and the behavior characteristics of the specific user; convert the time series into the frequency domain space, and identify the peaks and valleys and fluctuation patterns in the user power consumption curve;

[0043] A clustering module is used to determine the type of power load through the peaks, valleys and fluctuation patterns in the power consumption curve, and to adaptively cluster the power load type, wherein the characteristic vector of the adaptive clustering includes: peak information during the morning peak period and evening peak period of each day; valley information corresponding to all time points of each day; and the average value of the corresponding information of the daily user curve, thereby generating a suitable clustering area;

[0044] The control strategy generation module is used to generate a load control strategy that meets the needs of various users by combining the user classification results obtained by cluster analysis with other quantified key features. The other key features include geographical characteristics, industry characteristics, scale characteristics, load characteristics and load elasticity characteristics.

[0045] Further, including:

[0046] In the data preprocessing module, preprocessing the active power data includes determining whether the active power data of the user at the moment is abnormal according to the power change rate between adjacent or two fixed time points, and deleting the relevant data if it is abnormal data.

[0047] Further, including:

[0048] The determining whether the active power data of the user at the moment is abnormal according to the power change rate between adjacent or two fixed time points includes:

[0049] Traverse the collected historical active power data and corresponding time, and use the active power data at the same time in a certain period of time as a new data set, that is, form multiple data sets formed at fixed time points every day;

[0050] Taking each data set formed at the same fixed time point as the benchmark, the difference of active power data of adjacent data is calculated. If the difference between different days at the same time is not greater than the first threshold, the data are deemed valid; otherwise,

[0051] If the current difference is greater than the first threshold, the time corresponding to the two adjacent data is checked. If the data change is caused by weather reasons, the current two data are retained. Otherwise, the current data of the next day is deemed invalid, and the difference between the data of the next two days and the current data is calculated and iterative judgment is performed.

[0052] Taking each data set formed at different time points on the same day as a benchmark, the difference of active power data of adjacent data is calculated. If the difference calculated at different times on the same day is not greater than the second threshold, the data are deemed valid; otherwise,

[0053] If the current difference is greater than the second threshold, check the time corresponding to the two adjacent data. If the data change is caused by day and night, retain the current two data; otherwise, determine that the current data of the next day is an invalid value and delete it; continue to calculate the difference between the data of the next two days and the current data and then iterate and judge.

[0054] In a third aspect, the present invention provides an electronic terminal, comprising:

[0055] A memory, the memory being used to store a computer program;

[0056] A processor, wherein the processor is used to execute the computer program stored in the memory so that the electronic terminal executes the above-mentioned method for generating a load control strategy applicable to a regional power grid.

[0057] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the above-mentioned method for generating a load control strategy applicable to a regional power grid.

[0058] Beneficial effects: Compared with the prior art, the present invention has the following advantages:

[0059] (1) The present invention adopts two processing methods, horizontal and vertical, for preprocessing the collected historical active power data. The horizontal direction refers to different time points in the same day, and the vertical direction refers to the same time point in different days. According to different climates and the data characteristics of day and night in a day, the data is preprocessed in two directions to extract invalid data and retain valid data, thereby providing a basis for the implementation of the precise load control strategy below.

[0060] (2) The present invention captures the periodicity of the power consumption curve according to the time series, and dynamically identifies non-periodic fluctuation patterns through an adaptive algorithm to obtain a variety of user load types, and classifies users through a clustering algorithm, thereby ensuring accurate identification of user power consumption patterns, thereby providing important basic data for subsequent analysis and decision-making.

[0061] (3) The load characteristic analysis of the present invention not only takes into account the shape of the electricity consumption curve, but also deeply explores the user's electricity consumption behavior pattern, considering geographical characteristics, industry characteristics and scale characteristics as characteristic constraint inputs. Compared with the existing load control strategy generation method, it can achieve more accurate load control strategy generation, improve the pertinence and adaptability of load control, further reduce the operating cost of the power system, and improve the stability of the power supply system.

[0062] (4) The present invention quantifies the geographical characteristics of load users, the industry characteristics of load users, the scale characteristics of load users, the load characteristics of load users and the load elasticity characteristics, sets corresponding constraints, and uses a greedy algorithm to allocate electricity rights to each user, thereby providing a more accurate basis for load control and better meeting actual needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0064] Figure 1 A flow chart of a method for generating a load control strategy applicable to a regional power grid according to an embodiment of the present invention;

[0065] Figure 2 A schematic diagram of a daytime double-hump user curve of a load control strategy generation method applicable to a regional power grid according to an embodiment of the present invention;

[0066] Figure 3 A schematic diagram of an anti-peak user curve of a load control strategy generation method applicable to a regional power grid according to an embodiment of the present invention;

[0067] Figure 4 A schematic diagram of a continuous user curve of a method for generating a load control strategy applicable to a regional power grid according to an embodiment of the present invention;

[0068] Figure 5 A schematic diagram of the structure of an electronic device according to an embodiment of the present invention;

[0069] Figure 6 A schematic diagram of the structure of a load control strategy generation system applicable to a regional power grid according to an embodiment of the present invention;

[0070] Figure 7 A flow chart of a method for preprocessing a longitudinal data set according to an embodiment of the present invention;

[0071] Figure 8 This is a method for preprocessing a horizontal data set according to an embodiment of the present invention. DETAILED DESCRIPTION

[0072] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0073] Example 1

[0074] like Figure 1 As shown, in a first aspect of the present invention, the present invention provides a method for generating a load control strategy applicable to a regional power grid, the method comprising the following steps:

[0075] S1 obtains the historical active power data of the load user, and preprocesses the active power data to form a preprocessed power consumption data set, wherein the active power data includes data collected by a smart meter or a remote collection system.

[0076] Specifically, obtaining the historical active power data of load users includes:

[0077] Step (1.1): Obtain the historical active power data of load users, including data collected through smart meters or remote data collection systems.

[0078] Step (1.2): Continuously record and store the user's daily, hourly and minute electricity consumption, including high-frequency continuous collection and recording of power to ensure data continuity and integrity.

[0079] The collected power data is preprocessed to ensure the accuracy and reliability of the data, including the following steps:

[0080] Step (2.1): Data preprocessing, detect and remove possible duplicate data to avoid introducing redundant information during the analysis process. Ensure the consistency and correct format of timestamps for subsequent time-related analysis.

[0081] Step (2.2): Data change rate calculation: Calculate the power difference between adjacent time points to obtain the power change rate, thereby reducing noise and sudden fluctuations.

[0082] Step (2.3): Filter outliers. Define the reasonable range of power variation based on the power consumption of load users. Filter out data that is not within the range.

[0083] The specific method is: determine whether the active power data of the user at the moment is abnormal according to the power change rate between adjacent or two fixed time points, and if it is abnormal data, delete the relevant data.

[0084] Furthermore, this embodiment includes:

[0085] like Figure 7 As shown, the determining whether the active power data of the user at the moment is abnormal according to the power change rate between adjacent or two fixed time points includes:

[0086] The collected historical active power data and the corresponding time are traversed, and the active power data at the same time in a certain time period is used as a new data set, that is, multiple data sets formed at fixed time points every day are formed.

[0087] In this embodiment, if relevant data is collected every 15 minutes every day, and the collection time interval is at least two years, multiple data sets formed at fixed time points every day are formed, such as 2021.01.01:00:15+corresponding data A1; 2021.01.02:00:15+corresponding data A2; 2021.01.03:00:15+corresponding data A3, until the relevant data within two years forms the first data set, and then, 2021.01.01:00:30+corresponding data; 2 021.01.02:00:30+corresponding data; 2021.01.03:00:30+corresponding data, until the relevant data within two years form the second data set; 2021.01.01:00:45+corresponding data; 2021.01.02:00:45+corresponding data; 2021.01.03:00:45+corresponding data, until the relevant data within two years form the third data set, until the ninety-sixth data set is formed; the data set formed by this corresponding data set is a longitudinal data set.

[0088] Of course, the time involved in the above data may not be continuous, but a fixed interval time point, such as 2021.01.01:00:15+corresponding data A1; 2021.01.02:00:45+corresponding data; 2021.01.03:01:15+corresponding data A3, and the specific interval time is not restricted.

[0089] Taking each data set formed at the same fixed time point as the benchmark, the difference of active power data of adjacent data is calculated. If the difference between different days at the same time is not greater than the first threshold, the data are deemed valid; otherwise,

[0090] If the current difference is greater than the first threshold, the time corresponding to the two adjacent data is checked. If the data change is caused by temperature, the current two data are retained. Otherwise, the current data of the next day is considered invalid, and the difference between the data of the next two days and the current data is calculated and iterated judgment is performed. This process is repeated to obtain all the differences and all the valid data.

[0091] Specifically, in this embodiment, if the value of A2-A1 is calculated, if the absolute value of the corresponding value is not greater than the first threshold value set according to historical experience, the data is deemed valid, and the first threshold value α set according to historical experience is calculated by historical data that is not used this time. If A2-A1 is greater than the first threshold value, first confirm the time corresponding to the two rows of data, and confirm whether the weather changes are obvious. If the data changes are caused by weather differences, retain the current data as valid data. Otherwise, first determine A2 as invalid data, and use the above method to perform difference judgment on A3-A1 to confirm its validity.

[0092] This embodiment does not impose specific restrictions on whether the weather change is obvious, because different seasons, different places and different weather may cause an impact, and it is set according to actual conditions.

[0093] In this embodiment, a horizontal data set is also constructed, which is specifically: Figure 8 As shown, the data corresponding to different time points in the same day form corresponding data sets, such as 2021.01.01:00:15 + corresponding data A1; 2021.01.01:00:30 + corresponding data B1; 2021.01.01:00:45 + corresponding data C1, until the day and night related data in one day form the first data set; the related data within two years are formed into corresponding data sets, and the data finally obtained is recorded as the horizontal data set. In this embodiment, not only the data caused by climate is considered, but also the relationship between user load caused by day and night changes.

[0094] Taking each data set formed at different time points on the same day as a benchmark, calculate the difference in active power data of adjacent data. If the calculated difference at different times on the same day is not greater than the second threshold, then the data are deemed to be valid; that is, if the value of B1-A1 is not greater than the second threshold, then both data are deemed to be valid data.

[0095] Otherwise, if the current difference is greater than the second threshold, the time corresponding to the two adjacent data is checked. If the data change is caused by day and night, the current two data are retained. Otherwise, the current data of the next day is deemed to be an invalid value, and the difference between the data of the next two days and the current data is continued to be calculated and then iterative judgment is made.

[0096] Specifically, in this embodiment, if the absolute value corresponding to the value of B1-A1 is greater than the second threshold, it is queried whether the corresponding time is a data change caused by day and night changes, and different processing is performed according to different situations. Of course, this embodiment does not make specific restrictions on the data changes caused by specific day and night changes. At least there needs to be a time range limit, because each user's situation is different, which needs to be set according to the actual situation.

[0097] The embodiment of the present invention does not limit the specific values ​​of the first threshold and the second threshold, but in general, the change caused by climate is more obvious than the data change caused by day and night, so the first threshold is generally greater than the second threshold. However, this is not absolute, and some enterprises need to use a lot of electricity during the day or night, so the corresponding threshold will also increase.

[0098] Furthermore, the above method in this embodiment can also be trained using a neural network to make it more intelligent.

[0099] In addition, this embodiment also needs to correct some deleted data, such as:

[0100] Step (2.4): Interpolation data correction, identify and mark missing values ​​in the data. Use polynomial interpolation algorithm to fill in missing data to maintain data integrity.

[0101] S2 processes the preprocessed user power consumption data set in segments according to time stages and specific users, generates a time series, and dynamically adjusts it according to the time stages and the behavioral characteristics of the specific users;

[0102] S3 transforms the time series into frequency domain space and identifies peaks, valleys and fluctuation patterns in the user's power consumption curve.

[0103] In this embodiment, time domain and frequency domain analysis techniques are used to extract load characteristics to identify peaks, valleys and fluctuation patterns in user power consumption curves.

[0104] Sequence Analysis (SA): The user's historical electricity consumption dataset D is segmented to generate a time series S, and the analysis parameters are dynamically adjusted according to different time periods and specific user behavior characteristics to improve the analysis accuracy.

[0105] Frequency domain analysis (FA): The time series S is transformed into the frequency domain space F through adaptive Fourier transform, the frequency domain features are extracted, the periodicity of the power consumption curve is captured, and the non-periodic fluctuation patterns are dynamically identified through adaptive algorithms.

[0106] Through the joint analysis of time domain and frequency domain, a comprehensive feature vector FV is formed to reveal the peaks, valleys and fluctuation patterns in the power consumption curve:

[0107] FV=[SA(D),FA(SA(D))].

[0108] S4 determines the type of power load through the peaks, valleys and fluctuation patterns in the power consumption curve, and adaptively clusters the power load type. The characteristic vector of the adaptive clustering includes: peak information during the morning peak period and evening peak period of each day; valley information corresponding to all time points of each day; and the average value of the corresponding information of the daily user curve, thereby generating a suitable clustering area.

[0109] In this embodiment, Figure 2-4 As shown, the above data classification algorithm defines the data into three types of electricity load, namely, daytime double hump, anti-peak type and continuous type, so as to distinguish different electricity consumption behavior patterns and consumption habits.

[0110] Extract peak information from user load curves:

[0111] Daily peak value: P peak1 =max(P(t1,t2)),P peak2 =max(P(t3,t4));

[0112] Among them, t1 and t2 are morning peak hours, t3 and t4 are evening peak hours, and P is the active power value of the corresponding data.

[0113] P vallery =min(P(t));

[0114] Here, t represents all time points in a day.

[0115]

[0116] Where N is the total number of time points in a day.

[0117] Double hump during the day: The double hump user electricity consumption behavior during the day is manifested in that the load power consumption curve has obvious hump characteristics, that is, there are two power consumption peaks.

[0118] P peak1 ,P peak2 》P avg ;

[0119] Off-peak type: Off-peak users’ electricity consumption behavior is characterized by no obvious load during the day and a peak in production work plans at night.

[0120] P night 》P day ,P night >>P avg ;

[0121] If the load curve is relatively flat during the day but has an obvious hump at night, it is judged to be an anti-peak type.

[0122] Continuous type: The electricity consumption behavior of continuous users is relatively continuous, without obvious peaks.

[0123] max(P(t))-min(P(t))≈0.1×P avg ;

[0124] If the load curve fluctuates slightly within a day and has no obvious peaks or valleys, it is judged to be continuous.

[0125] The specific adaptive clustering algorithm can adopt the following scheme:

[0126] Define an adaptive clustering algorithm to classify user load feature vectors. Mainly includes:

[0127] Step (3.2.1): Construct feature vector: Extract the load feature of each user into a vector X = [P peak1 ,P peak2 ,P valley ,P avg ].

[0128] Step (3.2.2): Select clustering algorithm: Use clustering algorithms such as K-means to classify the feature vectors.

[0129] Step (3.2.3): Initialize cluster centers: Select K = 3 cluster centers, corresponding to the daytime double hump, reverse peak type and continuous type respectively.

[0130] Step (3.2.4.1): Calculate the distance from each data point to all cluster centers and assign each data point to the nearest cluster center. i ,C j )=||X i -C j ||

[0131] Among them, X i is the i-th data point, C j is the jth cluster center.

[0132] Step (3.2.4.2): Update the cluster center to the average value of all data points in the current cluster.

[0133]

[0134] Among them, S j is the set of data points in the jth cluster.

[0135] Based on the preliminary analysis results of the data, suitable clustering areas are automatically generated. Users are initially divided into three categories: daytime double hump, anti-peak type and continuous type. This classification method ensures accurate identification of user power consumption patterns, thereby providing important basic data for subsequent analysis and decision-making.

[0136] S5 combines the user classification results obtained by cluster analysis with other quantified key features to generate a load control strategy that meets the needs of various types of users. The other key features include geographical characteristics, industry characteristics, scale characteristics, load characteristics, and load elasticity characteristics.

[0137] In this embodiment, the load characteristic analysis not only takes into account the shape of the power consumption curve, but also deeply explores the power consumption behavior pattern of the user, providing a more accurate basis for load control and better meeting actual needs.

[0138] Based on the above cluster analysis results, step (4) further generates customized load control strategies for different load users through an optimized greedy algorithm. This step combines the user classification results obtained by cluster analysis with other key features (such as geographical features, industry characteristics, scale characteristics and load characteristics) to form the input parameters of the model and generates load control strategies that meet the needs of various types of users. It mainly includes the following steps:

[0139] Step (4.1): Analyze the characteristics of load users. Consider the following characteristics:

[0140] Geographical characteristics of load users G: The geographical characteristics are quantified. The spatial differences between load users are converted into constraints in the mathematical model to ensure that the load control strategy is more balanced in geographical distribution.

[0141] The location coordinates (such as longitude and latitude) of load users are quantified to form a geographic feature matrix.

[0142] Setting Constraints Among them, G i represents the geographical characteristic value of the i-th load user, x i represents the decision variable, i.e., whether to select the user to participate in load control, and K is the geographical distribution equilibrium constant.

[0143] Industry characteristics of load users I: In order to meet the needs of different industries, economic benefits are introduced as constraints. The electricity cost is compared with the output value to ensure that the generated load control strategy is more economically reasonable.

[0144] Calculate the ratio of electricity cost to output value of each industry to form an industry characteristic matrix. Introduce economic benefit constraints into the model.

[0145] Setting Constraints Among them, I i It represents the economic benefit value of the i-th load user, and M is the lower limit of the economic benefit.

[0146] Scale characteristics S of load users: Quantify users according to their electricity consumption scale. Through comprehensive evaluation of scale characteristics such as power consumption and power demand, users are divided into two main categories: large-scale enterprises and small-scale enterprises.

[0147] The scale characteristic constraint is introduced, and the selection of users with different load scales is achieved by setting different weight parameters.

[0148] Load characteristics L of load users: Considering the needs of the fast increase and decrease group and the load control group of the load control strategy, combined with the characteristics of load classification, set constraints to put the bi-peak users into the fast increase and decrease group. The load users in this group can achieve a rapid decrease in load in a short period of time, so as to formulate load control strategies more accurately and ensure that the load can be adjusted more flexibly during bi-peak power consumption.

[0149] Load elasticity feature E: Based on the user's historical response behavior, the user's load adjustment elasticity is calculated to form an elasticity feature matrix. Users with higher elasticity will be given priority in load control.

[0150] Step (4.2): Through multi-objective optimization, ensure that the generated load control strategy not only meets the needs of geographical balance and peak shifting, but also takes into account the actual conditions of large-scale and small-scale enterprises:

[0151] Step (4.3): U = (G, I, S, L, E)

[0152] Step (4.4): Then define a priority weight W for each user based on the features given in the above formula:

[0153] W=w1×G+w2×I+w3×S+w4×L+w5×E;

[0154] Among them, w1, w2, w3, w4, w5 are weight coefficients, which can be adjusted according to the importance of the features.

[0155] Consider i users, each user i has a feature vector U i and a weight W i In each time period t, select the user with the largest weight:

[0156]

[0157] Step (4.5): Allocate electricity usage rights to the user and update the load of the power grid and the weight of the user:

[0158] W i (t+1)=Wi (t)-δ;

[0159] Among them, δ is a positive number, representing the amount of weight reduction.

[0160] Step (4.6): Repeat the above process until all users are assigned electricity rights. Therefore, the final load control plan is determined by comprehensively analyzing the results of user electricity rights allocation.

[0161] Therefore, the user classification results obtained by cluster analysis are combined with other quantified key features to generate load control strategies that meet the needs of various users, including:

[0162] A corresponding feature vector is formed according to the geographical characteristics, industry characteristics, scale characteristics, load characteristics and load elasticity characteristics, and a priority weight is defined for the feature vector of each user; in each time period, the power usage right is allocated to the user with the largest selection weight, and the load of the power grid and the weight of the user are updated; this process is repeated until all users are allocated power usage rights, and the final load control plan is determined by comprehensively considering the results of the user power usage rights allocation.

[0163] Therefore, this embodiment can also use a greedy algorithm to allocate users, that is: the active power data of each user is divided into stages using a certain time period, the load type, feature vector and priority weight corresponding to each user are used as input, and a greedy algorithm is used to allocate electricity usage rights to each user.

[0164] Specifically: Use the greedy algorithm to allocate electricity rights to each user. In each period, select the user with the largest weight W to allocate electricity rights, then update the load of the power grid and the weight of the user, and repeat this process until all users are allocated electricity rights. Therefore, the final load control plan is determined by comprehensively analyzing the results of the allocation of user electricity rights.

[0165] Example 2

[0166] The present invention also provides a load control strategy generation system applicable to a regional power grid, the system comprising:

[0167] A data preprocessing module is used to obtain the historical active power data of the load user and preprocess the active power data to form a preprocessed power data set, wherein the active power data includes data collected by a smart meter or a remote collection system;

[0168] The power consumption curve recognition module is used to segment the pre-processed user power consumption data set according to the time stage and specific user, generate a time series, and dynamically adjust it according to the time stage and the behavior characteristics of the specific user; convert the time series into the frequency domain space, and identify the peaks and valleys and fluctuation patterns in the user power consumption curve;

[0169] A clustering module is used to determine the type of power load through the peaks, valleys and fluctuation patterns in the power consumption curve, and to adaptively cluster the power load type, wherein the characteristic vector of the adaptive clustering includes: peak information during the morning peak period and evening peak period of each day; valley information corresponding to all time points of each day; and the average value of the corresponding information of the daily user curve, thereby generating a suitable clustering area;

[0170] The control strategy generation module is used to generate a load control strategy that meets the needs of various users by combining the user classification results obtained by cluster analysis with other quantified key features. The other key features include geographical characteristics, industry characteristics, scale characteristics, load characteristics and load elasticity characteristics.

[0171] Further, including:

[0172] In the data preprocessing module, preprocessing the active power data includes determining whether the active power data of the user at the moment is abnormal according to the power change rate between adjacent or two fixed time points, and deleting the relevant data if it is abnormal data.

[0173] Further, including:

[0174] The determining whether the active power data of the user at the moment is abnormal according to the power change rate between adjacent or two fixed time points includes:

[0175] Traverse the collected historical active power data and corresponding time, and use the active power data at the same time in a certain period of time as a new data set, that is, form multiple data sets formed at fixed time points every day;

[0176] Taking each data set formed at the same fixed time point as the benchmark, the difference of active power data of adjacent data is calculated. If the difference between different days at the same time is not greater than the first threshold, the data are deemed valid; otherwise,

[0177] If the current difference is greater than the first threshold, the time corresponding to the two adjacent data is checked. If the data change is caused by temperature, the current two data are retained. Otherwise, the current data of the next day is deemed invalid, and the difference between the data of the next two days and the current data is calculated and iterative judgment is performed.

[0178] Taking each data set formed at different time points on the same day as a benchmark, the difference of active power data of adjacent data is calculated. If the difference calculated at different times on the same day is not greater than the second threshold, the data are deemed valid; otherwise,

[0179] If the current difference is greater than the second threshold, check the time corresponding to the two adjacent data. If the data change is caused by day and night, retain the current two data; otherwise, determine that the current data of the next day is an invalid value and delete it; continue to calculate the difference between the data of the next two days and the current data and iteratively judge that the first threshold is greater than the second threshold.

[0180] Other technical features of the load control strategy generation system applicable to a regional power grid described in the present invention are similar to the corresponding load control strategy generation method applicable to a regional power grid, and will not be repeated here.

[0181] Example 3

[0182] Figure 5 A schematic diagram of the structure of an electronic device 40 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0183] like Figure 5 As shown, the electronic device 40 includes at least one processor 41, and a memory connected to the at least one processor 41, such as a read-only memory (ROM) 42, a random access memory (RAM) 43, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 41 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 42 or the computer program loaded from the storage unit 48 to the random access memory (RAM) 43. In the RAM 43, various programs and data required for the operation of the electronic device 40 can also be stored. The processor 41, the ROM 42, and the RAM 43 are connected to each other through a bus 44. An input / output (I / O) interface 45 is also connected to the bus 44.

[0184] A number of components in the electronic device 40 are connected to the I / O interface 45, including: an input unit 46, such as a keyboard, a mouse, etc.; an output unit 47, such as various types of displays, speakers, etc.; a storage unit 48, such as a disk, an optical disk, etc.; and a communication unit 49, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 49 allows the electronic device 40 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0185] The processor 41 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 41 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 41 performs the various methods and processes described above.

[0186] In some embodiments, the load control strategy generation method applicable to the regional power grid may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 48. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 40 via the ROM 42 and / or the communication unit 49. When the computer program is loaded into the RAM 43 and executed by the processor 41, one or more steps of the load control strategy generation method applicable to the regional power grid described above may be performed. Alternatively, in other embodiments, the processor 41 may be configured to execute the load control strategy generation system applicable to the regional power grid in any other appropriate manner (e.g., by means of firmware).

[0187] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0188] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0189] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0190] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0191] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0192] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0193] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0194] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for generating a load control strategy for a regional power grid, characterized in that: The method includes: Obtaining historical active power data of load users, and preprocessing the active power data to form a preprocessed power data set, wherein the active power data includes data collected by a smart meter or a remote collection system; The pre-processed user power consumption data set is segmented according to time stages and specific users to generate a time series, and dynamically adjusted according to the time stages and the behavioral characteristics of the specific users; Convert the time series into frequency domain space and identify peaks, valleys and fluctuation patterns in the user's power consumption curve; Determine the type of power load by the peaks, valleys and fluctuation patterns in the power consumption curve, and adaptively cluster the power load type, and the characteristic vector of the adaptive clustering includes: peak information during the morning peak period and evening peak period of each day; valley information corresponding to all time points of each day; and the average value of the corresponding information of the daily user curve, thereby generating a suitable clustering area; The user classification results obtained by cluster analysis are combined with other quantified key features to generate load control strategies that meet the needs of various types of users. The other key features include geographical characteristics, industry characteristics, scale characteristics, load characteristics and load elasticity characteristics.

2. The method for generating a load control strategy applicable to a regional power grid according to claim 1, characterized in that: The acquisition of the load user's historical active power data includes: continuously recording and storing the user's daily, hourly, and minute power consumption, and high-frequency continuous collection and recording of power.

3. The method for generating a load control strategy applicable to a regional power grid according to claim 1, characterized in that: The preprocessing of the active power data includes determining whether the active power data of the user at the moment is abnormal according to the power change rate between adjacent or two fixed time points, and if it is abnormal data, deleting the relevant data.

4. The method for generating a load control strategy for a regional power grid according to claim 3, characterized in that: The determining whether the active power data of the user at the moment is abnormal according to the power change rate between adjacent or two fixed time points includes: Traverse the collected historical active power data and corresponding time, and use the active power data at the same time in a certain period of time as a new data set, that is, form multiple data sets formed at fixed time points every day; Taking each data set formed at the same fixed time point as the benchmark, the difference of active power data of adjacent data is calculated. If the difference between different days at the same time is not greater than the first threshold, the data are deemed valid; otherwise, If the current difference is greater than the first threshold, the time corresponding to the two adjacent data is checked. If the data change is caused by weather reasons, the current two data are retained. Otherwise, the current data of the next day is deemed invalid, and the difference between the data of the next two days and the current data is calculated and iterative judgment is performed. Taking each data set formed at different time points on the same day as a benchmark, the difference of active power data of adjacent data is calculated. If the difference calculated at different times on the same day is not greater than the second threshold, the data are deemed valid; otherwise, If the current difference is greater than the second threshold, the time corresponding to the two adjacent data is checked. If the data change is caused by day and night, the current two data are retained. Otherwise, the current data of the next day is deemed to be an invalid value, and the difference between the data of the next two days and the current data is continued to be calculated and then iterative judgment is made.

5. The method for generating a load control strategy applicable to a regional power grid according to claim 1, characterized in that: The identifying of peaks, valleys and fluctuation patterns in the user's electricity consumption curve includes: Dynamically identify non-periodic fluctuation patterns through adaptive algorithms; According to the peak and valley conditions of the user's electricity consumption curve, the peak information corresponding to the electricity consumption curve is extracted to determine three types of electricity load, including: daytime double hump, reverse peak type and continuous type. The daytime double hump refers to the user's electricity consumption behavior being manifested in that the load electricity consumption curve has an obvious hump feature, that is, there are two electricity consumption peaks; the reverse peak type refers to the user's electricity consumption behavior being manifested in that there is no obvious load during the day, but there is a peak of the production work plan at night; the continuous type refers to the user's electricity consumption behavior being relatively continuous, without an obvious peak, that is, the corresponding load curve fluctuates slightly within a day, and has no obvious peaks or troughs, and it is determined to be a continuous type.

6. The method for generating a load control strategy applicable to a regional power grid according to claim 5, characterized in that: The method of determining the type of power load by using the peaks, valleys and fluctuation patterns in the power consumption curve and adaptively clustering the power load type comprises: The relevant load characteristics of each user are extracted as a vector. The relevant load characteristics include: peak information during the morning peak period and the evening peak period; valley information corresponding to all time points of the day; and the average value of the corresponding information of the daily user curve; three clustering centers corresponding to daytime double hump, anti-peak type and continuous type are selected, and the clustering algorithm is used to classify the feature vectors; users are preliminarily divided into three categories: daytime double hump, anti-peak type and continuous type.

7. The method for generating a load control strategy for a regional power grid according to claim 6, characterized in that: The other key characteristics include: geographical characteristics, industry characteristics, scale characteristics, load characteristics and load elasticity characteristics, and the other key characteristics are quantified; The geographic feature G of the load user is processed as follows: the latitude and longitude position coordinates of the load user are quantized to form a geographic feature matrix, and a first constraint condition is set; The industry characteristics I of the load users are processed as follows: the ratio of electricity cost to output value of each industry is calculated to form an industry characteristics matrix, and a second constraint condition on economic benefits is introduced; The scale characteristic L of the load user is processed as follows: considering the requirements of the fast up and fast down group and the load control group of the load control strategy, setting the third constraint condition to put the bimodal users into the fast up and fast down group; The scale characteristics S of the load users are processed as follows: the users are quantified according to their electricity consumption scale, and the users are divided into two main categories: large-scale enterprises and small-scale enterprises according to their electricity consumption and electricity demand. The third constraint condition is introduced, and different weight parameters are set to select load users of different scales; The load elasticity characteristic E of the load user is processed as follows: the load adjustment elasticity of the user is calculated according to the historical response behavior of the user to form an elasticity characteristic matrix.

8. The method for generating a load control strategy applicable to a regional power grid according to claim 7, characterized in that: Combining the user classification results obtained by cluster analysis with other quantified key features, a load control strategy that meets the needs of various users is generated, including: A corresponding feature vector is formed according to the geographical characteristics, industry characteristics, scale characteristics, load characteristics and load elasticity characteristics, and a priority weight is defined for the feature vector of each user; in each time period, the power usage right is allocated to the user with the largest selection weight, and the load of the power grid and the weight of the user are updated; this process is repeated until all users are allocated power usage rights, and the final load control plan is determined by comprehensively considering the results of the user power usage rights allocation.

9. The method for generating a load control strategy applicable to a regional power grid according to claim 8, characterized in that: The active power data of each user is divided into stages according to a certain period of time. The load type, characteristic vector and priority weight corresponding to each user are taken as input, and the greedy algorithm is used to allocate electricity rights to each user.

10. A load control strategy generation system suitable for a regional power grid, characterized in that: The system includes: A data preprocessing module is used to obtain the historical active power data of the load user and preprocess the active power data to form a preprocessed power data set, wherein the active power data includes data collected by a smart meter or a remote collection system; The power consumption curve recognition module is used to segment the pre-processed user power consumption data set according to the time stage and specific user, generate a time series, and dynamically adjust it according to the time stage and the behavior characteristics of the specific user; convert the time series into the frequency domain space, and identify the peaks and valleys and fluctuation patterns in the user power consumption curve; A clustering module is used to determine the type of power load through the peaks, valleys and fluctuation patterns in the power consumption curve, and to adaptively cluster the power load type, wherein the characteristic vector of the adaptive clustering includes: peak information during the morning peak period and evening peak period of each day; valley information corresponding to all time points of each day; and the average value of the corresponding information of the daily user curve, thereby generating a suitable clustering area; The control strategy generation module is used to generate a load control strategy that meets the needs of various users by combining the user classification results obtained by cluster analysis with other quantified key features. The other key features include geographical characteristics, industry characteristics, scale characteristics, load characteristics and load elasticity characteristics.

11. The load control strategy generation system applicable to a regional power grid according to claim 10, characterized in that: In the data preprocessing module, preprocessing the active power data includes determining whether the active power data of the user at the moment is abnormal according to the power change rate between adjacent or two fixed time points, and deleting the relevant data if it is abnormal data.

12. The load control strategy generation system applicable to a regional power grid according to claim 11, characterized in that: The determining whether the active power data of the user at the moment is abnormal according to the power change rate between adjacent or two fixed time points includes: Traverse the collected historical active power data and corresponding time, and use the active power data at the same time in a certain period of time as a new data set, that is, form multiple data sets formed at fixed time points every day; Taking each data set formed at the same fixed time point as the benchmark, the difference of active power data of adjacent data is calculated. If the difference between different days at the same time is not greater than the first threshold, the data are deemed valid; otherwise, If the current difference is greater than the first threshold, the time corresponding to the two adjacent data is checked. If the data change is caused by weather reasons, the current two data are retained. Otherwise, the current data of the next day is deemed invalid, and the difference between the data of the next two days and the current data is calculated and iterative judgment is performed. Taking each data set formed at different time points on the same day as a benchmark, the difference of active power data of adjacent data is calculated. If the difference calculated at different times on the same day is not greater than the second threshold, the data are deemed valid; otherwise, If the current difference is greater than the second threshold, check the time corresponding to the two adjacent data. If the data change is caused by day and night, retain the current two data; otherwise, determine that the current data of the next day is an invalid value and delete it; continue to calculate the difference between the data of the next two days and the current data and then iterate and judge.

13. An electronic terminal, characterized in that: include: A memory, the memory being used to store a computer program; A processor, wherein the processor is used to execute the computer program stored in the memory so that the electronic terminal executes the load control strategy generation method applicable to the regional power grid as described in any one of claims 1-9.

14. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method for generating a load control strategy applicable to a regional power grid as described in any one of claims 1 to 9 is implemented.

Images