An intelligent time-delay closing electric energy meter control platform and method

Abstract

The present invention belongs to the technical field of electric energy meters, and discloses an intelligent delayed closing electric energy meter control platform and method; the method includes: acquiring the power consumption data and power consumption environment data of the electric energy meter; constructing a multi-dimensional data space based on the power consumption data and power consumption environment data of the electric energy meter; constructing a two-dimensional grid evolution model based on the multi-dimensional data space; identifying the peak power consumption area in the two-dimensional grid evolution model; for the identified peak power consumption area, using an improved sphere optimization algorithm to determine the optimal delayed closing time and duration to achieve the purpose of peak shaving and valley filling; not only optimizing the power grid operation, but also providing users with better power consumption services.

Description

Technical Field

[0001] The present invention relates to the technical field of electric energy meters, and more specifically, to an intelligent delayed closing electric energy meter control platform and method. Background Art

[0002] With the rapid development and popularization of smart grids, as a key power consumption data acquisition device, the control technology of electric energy meters is also constantly updated; traditional electric energy meter control systems mainly rely on simple power consumption records and basic load management, but in the face of increasingly complex power consumption environments and diverse user needs, they appear inadequate.

[0003] Traditional control platforms are difficult to effectively process and analyze massive multi-dimensional power consumption data, resulting in the inability to fully explore the deep power consumption patterns and laws contained in the data; this makes it difficult to accurately predict power consumption peaks and take effective load management measures in a timely manner; secondly, when identifying peak power consumption areas, existing methods are often too simplistic and static, unable to accurately capture dynamic power consumption behaviors, especially power consumption fluctuations during holidays or special weather conditions; in addition, traditional delayed closing strategies lack flexibility and intelligence, and are difficult to accurately adjust according to real-time power consumption situations and historical data, often resulting in unnecessary power outages or missing key peak shaving opportunities; more importantly, existing technologies often focus too much on short-term effects when implementing peak shaving and valley filling, ignoring comprehensive considerations of various factors such as user experience and long-term impacts; this one-sided control strategy has a negative impact on some important power-consuming facilities, affecting the overall stability and reliability.

[0004] In view of this, the present invention proposes an intelligent delayed closing electric energy meter control platform and method to solve the above problems. Summary of the Invention

[0005] To overcome the above-mentioned defects of the prior art and to achieve the above object, the present invention provides the following technical solution: An intelligent delayed closing electric energy meter control platform, comprising: a data acquisition module for acquiring power consumption data and power consumption environment data of the electric energy meter;

[0006] A space construction module for constructing a multi-dimensional data space based on the power consumption data and power consumption environment data of the electric energy meter;

[0007] A model construction module for constructing a two-dimensional grid evolution model based on the multi-dimensional data space;

[0008] An identification module for identifying peak power consumption areas in the two-dimensional grid evolution model;

[0009] An optimization control module is used to determine the optimal delay closing time and duration for the identified peak electricity consumption areas by using an improved sphere optimization algorithm; each module is connected by wired and / or wireless means.

[0010] Further, the electricity consumption data includes reading data, electricity consumption power curve, electricity consumption load type, electricity consumption equipment type, and user type; the electricity consumption environment data includes meteorological data, holiday information, electricity consumption area information, and grid load level.

[0011] Further, the method for constructing the multi-dimensional data space includes:

[0012] Clean and standardize the electricity consumption data and electricity consumption environment data of the electricity meter, perform numerical encoding on the categorical data to obtain the preprocessed electricity consumption data and electricity consumption environment data; merge the preprocessed electricity consumption data and electricity consumption environment data into a dataset X; each row in the dataset represents a sample; construct the samples into vector representations, denoted as sample vectors.

[0013] For each sample x in the dataset X, calculate the distance between it and all other samples; based on the obtained distances, construct an n×n distance matrix D, where n is the number of samples in the dataset X, and the element D(i, j) in the distance matrix D represents the distance between the i-th sample and the j-th sample.

[0014] For each sample x, find the k samples in the distance matrix D that are closest to x, i.e., the k nearest neighbors of x; k is a preset value and is an integer greater than 1; form a set by the sample x and its k nearest neighbors, denoted as the nearest neighbor set N(x); construct an undirected weighted graph G=(V, E, W), where V is the set of all nodes in the undirected weighted graph, corresponding to all samples in the dataset X; E is the set of all edges in the undirected weighted graph; W is the set of edge weights; convert the undirected weighted graph G into a fuzzy geometric graph; perform optimization and dimensionality reduction processing on the fuzzy geometric graph to obtain the multi-dimensional data space.

[0015] Further, the method for finding the k samples in the distance matrix D that are closest to x includes:

[0016] For each sample x, initialize k empty lists, copy the row corresponding to the sample x in the distance matrix D to a temporary list T1, sort the temporary list T1 to obtain an ordered list in ascending order of distance, and divide the ordered list into k sub-lists of equal size.

[0017] Take the element with the minimum distance from each sub - list to form a minimum heap. Take out the element with the minimum distance from the minimum heap, add it to the ordered list S, insert the next element in the sub - list where the taken - out element is located into the minimum heap, and repeat until all sub - lists are merged into the ordered list S. In the ordered list S, take out the first k elements, and their corresponding samples are the k nearest neighbors of x.

[0018] Furthermore, the method of converting the undirected weighted graph G into a fuzzy geometric graph includes:

[0019] For each node in the undirected weighted graph G , calculate its comprehensive local density ;

[0020] ; where, is the set of nearest neighbors belonging to node , is the distance between node and node and node ; is a control parameter, and its value ranges in (0, 1); is to its th nearest neighbor distance, is the importance factor between node and node ;

[0021] Importance factor ; where, is the similarity score between node and node , is the connectivity score between node and node ; is the weight coefficient, and its value ranges in (0, 1);

[0022] ; where, is the sample vector corresponding to node , is the sample vector corresponding to node , is the vector dot - product operation;

[0023] If there is a shortest path between node and node in the undirected weighted graph G ; then ; If node and node are directly connected in the undirected weighted graph G, then ;

[0024] find all nodes with a comprehensive local density greater than to form a set, denoted as the large density set , for each node in , calculate the distance between and , and take the minimum value of the distances between all nodes in and ; ; ;

[0025] For each node , calculate a clustering factor , sort all nodes in descending order according to the value of the clustering factor, starting from the node with the largest value of the clustering factor, divide it and its nearest neighbors into a cluster, and then move to the next unassigned node with the largest value of the clustering factor, repeating until all nodes are divided into clusters;

[0026] For each cluster, regard it as a simple geometry; for each node belonging to the cluster, define its membership degree as the reciprocal of the distance between the corresponding node and the central node of the corresponding cluster; the combination of all simple geometries and the membership degrees of their corresponding nodes constitutes a fuzzy geometric graph Gg.

[0027] Furthermore, the method for performing the optimized dimensionality reduction processing includes:

[0028] Define a low-dimensional space and determine the target dimension d of the low-dimensional space; for each sample x in the data set X, project it as a data point into the d-dimensional space to obtain a d-dimensional vector as the initial position of the data point in the low-dimensional space; in the low-dimensional space, based on the initial position of the data point, construct a low-dimensional fuzzy simple geometric graph G_low;

[0029] Define the loss function ;

[0030] ; where is the membership degree of x belonging to the convex hull shape in the fuzzy geometric graph Gg, is the membership degree of x belonging to the convex hull shape s in the low-dimensional fuzzy simple geometric graph G_low; is the weight coefficient of the regularization term, is the neighbor distribution of x in the fuzzy geometric graph Gg, is the neighbor distribution of x in the low-dimensional fuzzy simple geometric graph G_low; Denote the KL divergence; based on the loss function, iteratively optimize the positions of the data points in the low-dimensional space to minimize the loss function. In each iterative optimization, based on the current positions of the data points, construct a low-dimensional fuzzy simple geometric graph, calculate the value of the current loss function, and perform backpropagation on the positions of each data point to obtain the gradients. Use an optimizer to update the positions of each data point according to the gradients; repeat until reaching the preset maximum number of iterations; the finally obtained low-dimensional fuzzy simple geometric graph is the multi-dimensional data space.

[0031] Further, the construction method of the two-dimensional grid evolution model includes:

[0032] Map the multi-dimensional data space onto a two-dimensional plane, construct an N×N grid of squares on the mapped two-dimensional plane, and each grid point in the grid of squares represents a local area in the multi-dimensional data space; randomly assign an initial state to each grid point, and the state is alive or dead; define the neighbors of each grid point, and the neighbors include 8 adjacent grid points in the up, down, left, right, and four diagonal directions;

[0033] Set the evolution rule for the grid of squares, and the evolution rule is: if a grid point with a state of alive has 2 to 3 neighbors with a state of alive around it, then it remains in the state of alive in the next generation; if a grid point with a state of alive has less than 2 or more than 3 neighbors with a state of alive around it, its state becomes dead in the next generation; if a grid point with a state of dead has exactly 3 neighbors with a state of alive around it, its state becomes alive in the next generation;

[0034] Define the length of the time step; starting from the initial state, iteratively update the states of all grid points according to the evolution rule. In each step of iteration, for each grid point, find its neighbors in the grid of squares and calculate their states, and according to the evolution rule, determine the state of the corresponding grid point in the next time step, and at the same time update the states of all grid points; and at each time step, record the number of grid points with a state of alive. This grid of squares with iteratively updated states is the two-dimensional grid evolution model.

[0035] Further, the identification method of the peak electricity consumption area includes:

[0036] Set a time window with a fixed length, slide the time window along the time axis, and move one time step each time; during the process of sliding the time window, calculate the activity of each grid point within the time window; the calculation formula for the activity is:

[0037] ; where, is the number of time steps within the time window, is the index of the time step, is the grid point At the time step state, is the number of live states among the 8 neighbors around the grid point at the time step ; is the neighborhood influence coefficient, with a value in the range of (0, 1); for the grid point activity; is the step control parameter;

[0038] Organize the activity levels of grid points within the time window into a matrix, denoted as the activity matrix, and apply Gaussian smoothing or mean filtering to the activity matrix; Define a high-activity threshold, traverse the activity matrix, and mark the continuous region composed of grid points with activity levels higher than the high-activity threshold as the potential peak region;

[0039] Convert the activity matrix into a binary image, and set the pixels corresponding to the potential peak region in the binary image to 1 and the others to 0; Obtain the preliminary binary image; Apply dilation operation and erosion operation to the preliminary binary image; Obtain the mid-section potential peak region;

[0040] During the process of sliding the time window, for adjacent time windows, calculate the overlap degree of the mid-section potential peak regions obtained from them; Define an overlap degree threshold; If the overlap degree of the current time window with a certain previous time window is higher than the overlap degree threshold, record the duration between the two time windows;

[0041] Set a minimum duration threshold mdu, and mark those mid-section potential peak regions with a duration exceeding mdu as stable regions; Set the tracking period and the average threshold, and for each pair of stable regions, calculate their average overlap degree during the entire tracking period. If the average overlap degree is higher than the average threshold, merge the corresponding two stable regions, that is, calculate the union of the corresponding two stable regions in all time windows, and use this union as the new merged region, and perform morphological processing on the new merged region to obtain the final electricity peak region.

[0042] Furthermore, the determination method of the optimal delayed closing time and duration includes:

[0043] Define decision variables, where the decision variables include the delayed closing time and the duration ; Define the objective function ;

[0044] ; where , , and are the objective weight coefficients, is the original power consumption at time after the implementation of the delayed closing, the power consumption at time represents the positive part, is the cumulative effect factor of the historical delay impact; is the user impact factor, is the load balancing factor;

[0045] ; where, is the power change rate weight coefficient; is the historical impact weight coefficient; is the power change amount at time is the time interval; is the historical delay time point; is the time decay constant;

[0046] ; where, is the importance weight coefficient of the th type of user, is the number of the th type of user at time is the th type of user's cumulative delay time at time;

[0047] ; where, is the load balancing coefficient; is the maximum load value during the observation period, after the implementation of the delayed closing, the actual load value at time is the minimum load value during the observation period;

[0048] Randomly generate N2 spheres, where N2 is an integer greater than 1. Each sphere contains a center position and a radius. The center position is composed of decision variables, i.e., (x1, x2). Sort all the spheres in descending order according to the function values of the objective function; for each sphere, randomly generate M candidate solutions on the sphere surface, calculate the function values of the objective function of the candidate solutions, take the candidate solution with the largest function value as the optimal candidate solution, update the center position of the sphere to the position of the optimal candidate solution, and adjust the radius of the sphere;

[0049] The formula for adjusting the radius of the sphere is: ; where, is the adjusted radius, is the radius of the sphere corresponding to the optimal candidate solution ; The sphere selected from all the spheres The sphere with the largest function value among the candidate solutions outside ; is a random perturbation factor, and its value range is (-0.1, 0.1);

[0050] Sort all the spheres, eliminate the spheres with smaller function values, and take the spheres with adjusted radii as the new spheres; Iteratively and repeatedly calculate the change rate of the optimal candidate solution. If the change rate is less than the preset change threshold for K consecutive times, then take the optimal candidate solution of the last time as the final solution, and extract the decision variables corresponding to the final solution to obtain the best delayed closing time and duration.

[0051] An intelligent delayed closing electric energy meter control method, which is implemented based on the described intelligent delayed closing electric energy meter control platform, includes: Step 1, obtain the power consumption data and power consumption environment data of the electric energy meter;

[0052] Step 2, construct a multi-dimensional data space based on the power consumption data and power consumption environment data of the electric energy meter;

[0053] Step 3, construct a two-dimensional grid evolution model based on the multi-dimensional data space;

[0054] Step 4, identify the peak power consumption area in the two-dimensional grid evolution model;

[0055] Step 5, for the identified peak power consumption area, use an improved sphere optimization algorithm to determine the best delayed closing time and duration.

[0056] The technical effects and advantages of the intelligent delayed closing electric energy meter control platform and method of the present invention:

[0057] The present invention can effectively improve the operation efficiency and stability of the power grid and achieve precise power consumption load management; By deeply mining complex power consumption data, it can accurately identify the dynamically changing peak power consumption areas, thereby realizing more flexible and precise peak shaving and valley filling control; This not only helps to balance the power grid load, but also significantly reduces the operating costs of the power system; At the same time, it fully considers the user experience, can formulate more user-friendly control strategies according to different user types and historical power consumption situations, and effectively reduce the impact on users' daily power consumption; In addition, its intelligent characteristics enable it to adapt to complex and changeable power consumption environments, improve the adaptive ability and reliability of power energy management. Generally speaking, it not only optimizes the power grid operation, but also provides users with better power consumption services. Brief Description of the Drawings

[0058] Figure 1 It is a schematic diagram of an intelligent delayed closing electric energy meter control platform of the present invention;

[0059] Figure 2 Schematic diagram of an intelligent delayed closing electric energy meter control method of the present invention. Specific implementation manners

[0060] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0061] Embodiment 1

[0062] Please refer to Figure 1 As shown, an intelligent delayed closing electric energy meter control platform in this embodiment includes: a data acquisition module for acquiring the electricity consumption data and electricity consumption environment data of the electric energy meter;

[0063] A space construction module for constructing a multi-dimensional data space based on the electricity consumption data and electricity consumption environment data of the electric energy meter;

[0064] A model construction module for constructing a two-dimensional grid evolution model based on the multi-dimensional data space;

[0065] An identification module for identifying the electricity consumption peak area in the two-dimensional grid evolution model;

[0066] An optimization control module for determining the optimal delayed closing time and duration for the identified electricity consumption peak area by using an improved sphere optimization algorithm to achieve the purpose of peak shaving and valley filling; each module is connected by wired and / or wireless means to realize data transmission between modules.

[0067] The electricity consumption data includes reading data, electricity consumption power curve, electricity consumption load type (numerical coding), electricity consumption equipment type (numerical coding), and user type (numerical coding); the reading data includes total electricity consumption (degree) and sub-period electricity consumption (degree / period); electricity consumption power (kilowatt / time point), reflecting the change of electricity consumption power with time; electricity consumption load type, such as residential electricity consumption, industrial electricity consumption, commercial electricity consumption, etc.; electricity consumption equipment type, such as air conditioner, lighting, electrical appliances, etc.; user type, such as household user, enterprise user, etc.; the reading data and electricity consumption power curve are sequential data, changing with time. For this type of data, numerical coding is used, such as residential electricity consumption is 1, industrial electricity consumption is 2, etc.

[0068] The electricity consumption environment data includes meteorological data, holiday information (0 / 1 coding), electricity consumption area information (numerical coding), and grid load level (percentage); the meteorological data includes temperature, humidity, and sunshine duration, which are environmental factors affecting electricity consumption; electricity consumption area information, such as city, region, etc.

[0069] Directly processing the original high-dimensional data will face the curse of dimensionality, with high computational complexity, low efficiency, and simple dimensionality reduction methods (such as principal component analysis) may lose important non-linear relationships and local structural information. In the original high-dimensional space, the distribution and relationships of the data may not be obvious, making it difficult to discover meaningful electricity consumption patterns and rules.

[0070] The ways to construct a multi-dimensional data space include:

[0071] Clean and standardize the electricity consumption data and electricity consumption environment data of the electricity meter, dealing with missing values, outliers, etc.; numerically encode categorical data (such as electricity load type, electricity equipment type, user type, electricity consumption area information, etc.) to obtain the preprocessed electricity consumption data and electricity consumption environment data;

[0072] Merge the preprocessed electricity consumption data and electricity consumption environment data into a dataset X; each row in the dataset represents a sample, containing all electricity consumption data (such as reading data, electricity power curve, etc.) and electricity consumption environment data (such as meteorological data, holiday information, etc.); construct the samples into vector representations, denoted as sample vectors.

[0073] For each sample x in the dataset X, calculate the distance between it and all other samples. The distance calculation is the Euclidean distance, Manhattan distance, or cosine similarity between the corresponding sample vectors of the samples.

[0074] Based on the obtained distances, construct an n×n distance matrix D, where n is the number of samples in the dataset X, and the element D(i, j) in the distance matrix D represents the distance between the i-th sample and the j-th sample.

[0075] For each sample x, find the k samples in the distance matrix D that are closest to x, that is, the k nearest neighbors of x; k is a preset value and is an integer greater than 1; specifically, for each sample x, initialize k empty lists, copy the row in the distance matrix D corresponding to sample x (representing the distances between sample x and all other samples) to a temporary list T1, sort the temporary list T1 to obtain an ordered list in ascending order of distance, and divide the ordered list into k sub-lists of equal size.

[0076] Take the element with the smallest distance from each sub-list to form a minimum heap, take the element with the smallest distance from the minimum heap, add it to the ordered list S, insert the next element in the sub-list where the taken element is located into the minimum heap, repeat until all sub-lists are merged into the ordered list S. In the ordered list S, take the first k elements, and the samples corresponding to them are the k nearest neighbors of x; form a set with sample x and its k nearest neighbors, denoted as the nearest neighbor set N(x).

[0077] Construct an undirected weighted graph \(G=(V, E, W)\), where \(V\) is the set of all nodes in the undirected weighted graph, corresponding to all samples (nodes) in the dataset \(X\); \(E\) is the set of all edges in the undirected weighted graph, that is, if sample \(x\) and sample \(y\) are each other's nearest neighbors, then an undirected edge \(l(x, y)\) is added between \(x\) and \(y\); \(W\) is the set of edge weights, and the element \(W(x, y)\) in it represents the weight of the edge \(l(x, y)\), which is set to the distance between \(x\) and \(y\); Convert the undirected weighted graph \(G\) into a fuzzy geometric graph.

[0078] Specifically, for each node in the undirected weighted graph \(G\) , calculate its comprehensive local density ;

[0079] ; where is the set of nearest neighbors belonging to node , is the node within is node and node The distance between them (similarity between sample vectors); is a control parameter, with a value in the range \((0, 1)\); is to its The distance to the nearest neighbor (excluding ), is node and node The importance factor between them.

[0080] Importance factor ; where is the similarity score between node and node , is the connectivity score between node and node ; is the weight coefficient, used to balance the weights of the two items, with a value in the range \((0, 1)\).

[0081] ; where is the sample vector corresponding to node , is the sample vector corresponding to node , is the vector dot product operation; if there is a shortest path between node and node in the undirected weighted graph \(G\) ; then ; If node and node are directly connected in the undirected weighted graph G, then .

[0082] Find all nodes with a comprehensive local density greater than , form a set, denoted as the large density set . For each node in , calculate the distance between and , and take the minimum value of the distances between all nodes in and . .

[0083] For each node , calculate a clustering factor . Sort all nodes in descending order according to the value of the clustering factor. Starting from the node with the largest value of the clustering factor, divide it and its nearest neighbors into a cluster, and then move to the next node with the largest value of the unpartitioned clustering factor, repeating until all nodes are divided into clusters.

[0084] For each cluster, regard it as a simple geometry; for each node belonging to the cluster, define its membership degree as the reciprocal of the distance between the corresponding node and the central node of the corresponding cluster; the combination of all simple geometries and the membership degrees of their corresponding nodes constitutes a fuzzy geometric graph Gg; where each data point is assigned a membership degree belonging to a certain simple geometry; this representation preserves the local and global topological structures of the data.

[0085] Perform optimization and dimensionality reduction on the fuzzy geometric graph. Specifically, define a low-dimensional space and determine the target dimension d of the low-dimensional space. Usually, d = 2 or 3 for visualization; for each sample x in the dataset X, take it as a data point and project it into the d-dimensional space to obtain a d-dimensional vector as the initial position of the data point in the low-dimensional space; the projection can be a random projection method.

[0086] In the low-dimensional space, based on the initial positions of the data points, construct a low-dimensional fuzzy simple geometric graph G_low, and the calculation method is similar to that of the fuzzy geometric graph; define the loss function ;

[0087] ; where is the membership degree of x belonging to the convex hull in the fuzzy geometric graph Gg, and is the membership degree of x belonging to the convex hull s in the low-dimensional fuzzy simple geometric graph G_low; is the weight coefficient of the regular term, is the neighbor distribution of x in the fuzzy geometric graph Gg, is the neighbor distribution of x in the low-dimensional fuzzy simple geometric graph G_low; represents the KL divergence; it should be noted that the convex hull is a basic geometric structure in a geometric graph, used to represent the proximity relationship and topological structure between data points. Specifically, a J-simplex is a convex hull formed by J + 1 linearly independent points (vertices); for example, a 0-simplex is a single point, a 1-simplex is a line segment composed of 2 vertices, a 2-simplex is a triangle composed of 3 vertices, a 3-simplex is a tetrahedron composed of 4 vertices, and so on; the neighbor distribution is represented by a probability distribution, and each value in the probability distribution corresponds to the reciprocal of the distance from x to another node in the cluster.

[0088] Based on the loss function, the positions of the data points in the low-dimensional space are iteratively optimized to minimize the loss function. In each iterative optimization, based on the current positions of the data points, a low-dimensional fuzzy simple geometric graph is constructed, the value of the current loss function is calculated, and the positions of each data point are backpropagated to obtain the gradients. The optimizer (such as Adam) is used to update the positions of each data point according to the gradients; repeat until the preset maximum number of iterations is reached; the finally obtained low-dimensional fuzzy simple geometric graph is the multi-dimensional data space.

[0089] By minimizing the cross-entropy loss function, the low-dimensional fuzzy simple geometric graph in the low-dimensional space is forced to be as close as possible to G_high in the high-dimensional space; since G_high retains the local and global topological structures of the original data, the optimized low-dimensional fuzzy simple geometric graph will also retain these structural information as much as possible; specifically, if two samples are neighbors (belong to the same convex hull) in the high-dimensional space, then in the low-dimensional space, their positions will also be pulled closer; conversely, if they are not neighbors in the high-dimensional space, then the distance in the low-dimensional space will be pulled apart.

[0090] The power consumption data and power consumption environment data of the electricity meter are projected into a secure space, thereby constructing a multi-dimensional data space; in this multi-dimensional data space, similar samples are close to each other, while different samples are far apart, well retaining the essential structure and characteristics of the original high-dimensional data.

[0091] The construction method of the two-dimensional grid evolution model includes:

[0092] Map the multi-dimensional data space onto a two-dimensional plane, for example, using algorithms such as t-SNE. Construct an N×N grid of squares on the mapped two-dimensional plane. Each grid point in the grid of squares represents a local area in the multi-dimensional data space, and the distance between grid points reflects the similarity of data points in the original multi-dimensional data space. Randomly assign an initial state to each grid point, where the state is either alive (1) or dead (0). Define the neighbors of each grid point, and the neighbors include 8 adjacent grid points in the up, down, left, right, and four diagonal directions. For grid points on the boundary of the grid of squares, connect their opposite sides to form a toroidal structure, meaning that the left boundary of the grid is connected to the right boundary, and the upper boundary is connected to the lower boundary.

[0093] Set the evolution rules for the grid of squares. The evolution rules are as follows: If a grid point with a state of alive has 2 to 3 neighbors with a state of alive around it, then it remains in the state of alive in the next generation. If a grid point with a state of alive has less than 2 or more than 3 neighbors with a state of alive around it, its state becomes dead in the next generation. If a grid point with a state of dead has exactly 3 neighbors with a state of alive around it, its state becomes alive in the next generation.

[0094] Define the length of the time step and determine the actual time length represented by each state update, such as 15 minutes or 1 hour. Starting from the initial state, iteratively update the states of all grid points according to the evolution rules. In each step of iteration, for each grid point, find its neighbors in the grid of squares and calculate their states. According to the evolution rules, determine the state of the corresponding grid point in the next time step, and at the same time update the states of all grid points. And at each time step, record the number of grid points with a state of alive. This grid of squares with iterative state updates is the two-dimensional grid evolution model. The two-dimensional grid evolution model can better capture the complex relationships and structures of the original high-dimensional data and can more accurately simulate and predict complex electricity consumption behavior patterns.

[0095] The identification methods for peak electricity consumption areas include:

[0096] Set a time window with a fixed length, such as 4 hours or 8 hours, to observe the state changes of the two-dimensional grid evolution model. Slide the time window along the time axis, moving one time step length (such as 15 minutes or 1 hour) each time. During the process of sliding the time window, calculate the activity of each grid point within the time window. The calculation formula for activity is:

[0097] ; where is the number of time step lengths within the time window, is the index of the time step length, is the grid point at the time step length state (1 represents alive, 0 represents dead), is at the time step length Time lattice point The number of live states among the 8 neighbors around it, is the neighborhood influence coefficient, with a value in the range of (0, 1), used to balance local and global information, that is, to balance between considering the state of a single lattice point and the state of its neighborhood; is the lattice point activity; is the step control parameter, with a value in the range of (0, 1), used to regulate the balance of the distance between steps.

[0098] Organize the activity of lattice points within the time window into a matrix, denoted as the activity matrix, and apply Gaussian smoothing or mean filtering to the activity matrix to reduce noise and highlight the main features; define a high activity threshold (such as 0.8), traverse the activity matrix, and mark the continuous area composed of lattice points with activity higher than the high activity threshold as the potential peak area.

[0099] Convert the activity matrix into a binary image, and set the pixels corresponding to the potential peak area on the binary image to 1 and the others to 0; obtain the preliminary binary image; apply dilation operation and erosion operation to the preliminary binary image to merge adjacent areas and remove small isolated areas; obtain the mid-section potential peak area.

[0100] During the process of sliding the time window, for adjacent time windows, calculate the overlap degree of the mid-section potential peak areas obtained by them; the overlap degree is measured by the Jaccard coefficient; define an overlap degree threshold (such as 0.6) to judge whether the area persists; if the overlap degree of the current time window with a certain previous time window is higher than the overlap degree threshold, it is considered a continuation of the same area; then record the duration between the two time windows; if it is not higher than the overlap degree threshold, it is regarded as a newly emerging area, and update the duration of each area.

[0101] Set a minimum duration threshold mdu (such as 3 time windows), and mark those mid-section potential peak areas with a duration exceeding mdu as stable areas; set the tracking period and average threshold, for each pair of stable areas, calculate their average overlap degree during the entire tracking period, if the average overlap degree is higher than the average threshold, merge the corresponding two stable areas, that is, calculate the union of the corresponding two stable areas in all time windows, and use this union as the new merged area, and perform morphological processing (such as dilation or closing operation) on the new merged area to smooth the boundary and obtain the final electricity peak area.

[0102] The determination methods of the optimal delayed closing time and duration include:

[0103] Define decision variables, and the decision variables include the delayed closing time (Minute) and duration (Minute); define the objective function ;

[0104] ; where , , and are the objective weight coefficients, is the original power consumption at time is the power consumption at time after implementing the delayed closing, represents the positive part, that is, only the reduction amount is counted, is the cumulative effect factor of the historical delay impact; is the user impact factor, is the load balance factor.

[0105] ; where, is the power change rate weight coefficient, used to control the importance of the power change speed, and the value range is (0, 1); is the historical impact weight coefficient, used to control the importance of the historical delay impact, and the value range is (0, 1); is the power change amount at time is the time interval, usually taken as 15 minutes or 1 hour; is the historical delay time point; is the time decay constant, which controls the decay speed of the historical impact, and the unit is hour.

[0106] ; where, is the importance weight coefficient of the th type of user, and the value range is (0, 1); for example, the weight of important industrial users is relatively large (0.8 - 1.0), and the weight of ordinary residential users is relatively small (0.3 - 0.5); is the number of the th type of user at time is the th type of user's cumulative delay time at time in hours.

[0107] ; where, is the load balance coefficient, used to adjust the importance of the valley filling effect, and the value range is (0, 1); is the maximum load value during the observation period, is the power consumption at time The actual load value at a moment, is the minimum load value during the observation period.

[0108] Randomly generate N2 spheres, where N2 is an integer greater than 1. Each sphere contains a center position and a radius. The center position is composed of decision variables, i.e., (x1, x2). Sort all the spheres in descending order according to the function values of the objective function; for each sphere, randomly generate M candidate solutions (center positions) on the sphere surface, calculate the function values of the objective function of the candidate solutions, take the candidate solution with the largest function value as the optimal candidate solution, update the center position of the sphere to the position of the optimal candidate solution, and adjust the radius of the sphere; the formula for adjusting the radius of the sphere is: ; where, is the adjusted radius, is the radius of the sphere corresponding to the optimal candidate solution ; is the sphere selected from all the spheres and is the sphere with the largest function value among the candidate solutions outside the sphere ; is a random perturbation factor, and its value range is (-0.1, 0.1). In order to increase diversity and avoid completely identical search ranges.

[0109] Sort all the spheres, eliminate the spheres with smaller function values, and take the spheres with adjusted radii as the new spheres; iteratively and repeatedly calculate the change rate of the optimal candidate solution, mainly considering the changes in function values, the change in the delayed closing time (x1), and the change in the duration (x2); comprehensively consider the changes in these three aspects through a weighted method; if the change rate is less than the preset change threshold for K consecutive times, then take the last optimal candidate solution as the final solution, extract the decision variables corresponding to the final solution, and obtain the best delayed closing time and duration.

[0110] This embodiment can effectively improve the operation efficiency and stability of the power grid and achieve precise electricity load management; by deeply mining complex electricity consumption data, it can accurately identify dynamically changing electricity peak regions, thereby achieving more flexible and precise peak shaving and valley filling control; this not only helps to balance the power grid load, but also significantly reduces the operating costs of the power system; at the same time, it fully considers the user experience, can formulate more user-friendly control strategies according to different user types and historical electricity consumption situations, and effectively reduce the impact on users' daily electricity consumption; in addition, its intelligent characteristics enable it to adapt to complex and changeable electricity consumption environments, improve the adaptive ability and reliability of power energy management. Generally speaking, it not only optimizes the operation of the power grid, but also provides users with better power consumption services.

[0111] Embodiment 2

[0112] Please refer to Figure 2As shown in the figure, for the parts not described in detail in this embodiment, refer to the description in Embodiment 1. A method for controlling an intelligent delayed closing watt-hour meter is provided, including:

[0113] Step 1: Obtain the power consumption data and power consumption environment data of the watt-hour meter;

[0114] Step 2: Based on the power consumption data and power consumption environment data of the watt-hour meter, construct a multi-dimensional data space;

[0115] Step 3: Based on the multi-dimensional data space, construct a two-dimensional grid evolution model;

[0116] Step 4: Identify the peak power consumption area in the two-dimensional grid evolution model;

[0117] Step 5: For the identified peak power consumption area, use an improved sphere optimization algorithm to determine the optimal delayed closing time and duration.

[0118] Embodiment 3

[0119] This embodiment publicly provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, it implements the operation mode of the above-provided method for controlling an intelligent delayed closing watt-hour meter.

[0120] Since the electronic device introduced in this embodiment is the electronic device used to implement the method for controlling an intelligent delayed closing watt-hour meter in the embodiments of the present application, based on the method for controlling an intelligent delayed closing watt-hour meter introduced in the embodiments of the present application, those skilled in the art can understand the specific implementation manners and various variations of the electronic device in this embodiment. Therefore, the specific implementation of how this electronic device implements the method in the embodiments of the present application will not be described in detail here. As long as those skilled in the art implement the electronic device used in the method for controlling an intelligent delayed closing watt-hour meter in the embodiments of the present application, it falls within the scope of protection of the present application.

[0121] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain a formula that is closest to the actual situation. The preset parameters and threshold selection in the formulas are set by those skilled in the art according to the actual situation.

[0122] The above is only the preferred embodiment of the present invention. The protection scope of the present invention is not limited to the above embodiments. Any technical solutions falling within the concept of the present invention belong to the protection scope of the present invention. It should be noted that for those ordinary technical users in the technical field, several improvements and refinements made without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. An intelligent time-delay closing electric energy meter control platform, characterized in that: include: The data acquisition module is used to obtain the electricity consumption data and electricity consumption environment data of the electric energy meter; the electricity consumption environment data includes meteorological data, holiday information, electricity consumption area information and grid load level; The space construction module constructs a multi-dimensional data space based on the electricity consumption data and electricity consumption environment data of the energy meter; The model building module builds a two-dimensional grid evolution model based on the multi-dimensional data space; The construction method of the two-dimensional grid evolution model includes: The multidimensional data space is mapped onto a two-dimensional plane, and an N×N square grid is constructed on the mapped two-dimensional plane. Each grid point in the square grid represents a local area in the multidimensional data space. Each grid point is randomly assigned an initial state, which is either alive or dead. For each grid point, its neighbors are defined, and the neighbors include 8 adjacent grid points in the up, down, left, right, and four diagonal directions. An evolution rule is set for the square grid, and starting from the initial state, the states of all grid points are iteratively updated according to the evolution rule. The identification module is used to identify the peak power consumption area in the two-dimensional grid evolution model; the identification method of the peak power consumption area includes: Set a time window of fixed length, slide the time window along the time axis, and move one time step each time; in the process of sliding the time window, calculate the activity of each grid point in the time window; organize the activity of the grid points in the time window into a matrix, recorded as the activity matrix, and apply Gaussian smoothing or mean filtering to the activity matrix; define a high activity threshold, traverse the activity matrix, and mark the continuous area composed of grid points with activity higher than the high activity threshold as a potential peak area; in the process of sliding the time window, for adjacent time windows, calculate the overlap of their obtained middle potential peak areas; The optimization control module is used to determine the optimal delay closing time and duration for the identified power consumption peak area by using an improved sphere optimization algorithm; the optimal delay closing time and duration are determined by: Define decision variables, including delayed closing time x1 and duration x2; define objective function f(x1,x2); f(x1,x2)=a1×∑[P(t')-P'(t')] + -a2×∑D(t')-a3×C(t')+ a4×E(t'); where a1, a2, a3 and a4 are target weight coefficients, P(t') is the original power consumption at time t', P'(t') is the power consumption at time t' after delayed closing, [ ] + represents the positive part, D(t') is the cumulative effect factor of historical delay impact; C(t') is the user impact factor, and E(t') is the load balancing factor; N2 spheres are randomly generated, N2 is an integer greater than 1, each sphere contains a center position and a radius, and the center position is composed of decision variables, that is, (x1, x2), and all spheres are sorted in descending order according to the function value of the objective function; for each sphere, M candidate solutions are randomly generated on the surface of the sphere, and the function value of the objective function of the candidate solution is calculated. The candidate solution with the largest function value is taken as the optimal candidate solution, the center position of the sphere is updated to the position of the optimal candidate solution, and the radius of the sphere is adjusted; each module is connected by wired and / or wireless means.

2. The intelligent time-delay closing electric energy meter control platform according to claim 1 is characterized in that: The electricity consumption data includes reading data, electricity consumption power curve, electricity load type, electricity consumption equipment type and user type.

3. The intelligent time-delay closing electric energy meter control platform according to claim 2 is characterized in that: The method of constructing the multidimensional data space includes: The electricity consumption data and electricity consumption environment data of the electric energy meter are cleaned and standardized, and the categorical data are numerically encoded to obtain the preprocessed electricity consumption data and electricity consumption environment data; the preprocessed electricity consumption data and electricity consumption environment data are merged into a data set X; each row in the data set represents a sample; the sample is constructed into a vector representation, which is recorded as a sample vector; For each sample x in the data set X, calculate the distance between it and all other samples; based on the obtained distance, construct an n×n distance matrix D, where n is the number of samples in the data set X, and the element D(i, j) in the distance matrix D represents the distance between the i-th sample and the j-th sample; For each sample x, find the k samples closest to x in the distance matrix D, that is, the k nearest neighbors of x; k is a preset value and an integer greater than 1; the sample x and its k nearest neighbors form a set, recorded as the nearest neighbor set N(x); construct an undirected weighted graph G = (V, E, W), where V is the set of all nodes in the undirected weighted graph, corresponding to all samples in the data set X; E is the set of all edges in the undirected weighted graph; W is the set of edge weights; convert the undirected weighted graph G into a fuzzy geometric graph; optimize the fuzzy geometric graph for dimensionality reduction to obtain a multidimensional data space.

4. The intelligent time-delay closing electric energy meter control platform according to claim 3 is characterized in that: The method of finding the k samples closest to x in the distance matrix D includes: For each sample x, initialize k empty lists, copy the row corresponding to sample x in the distance matrix D to a temporary list T1, sort the temporary list T1 to obtain an ordered list in increasing distance, and divide the ordered list into k sublists of equal size; Take out the element with the smallest distance from each sublist to form a minimum heap, take out the element with the smallest distance from the minimum heap, add it to the ordered list S, insert the next element in the sublist where the taken element is located into the minimum heap, and repeat until all sublists are merged into the ordered list S. In the ordered list S, take out the first k elements, and their corresponding samples are the k nearest neighbors of x.

5. The intelligent time-delay closing electric energy meter control platform according to claim 4 is characterized in that: The method of converting the undirected weighted graph G into a fuzzy geometric graph includes: For each node x' in the undirected weighted graph G, calculate its comprehensive local density rho(x'); Where y′ is a node in the nearest neighbor set N(x′) of node x′, d(x′,y′) is the distance between node x′ and node y′; c is a control parameter with a value in the range of (0, 1); d_k(x′) is the distance from x′ to its kth nearest neighbor, and imp(x′,y′) is the importance factor between node x′ and node y′; Importance factor imp(x',y') = w1×sim(x',y')+(1-w1)×conn(x',y'); where sim(x',y') is the similarity score between node x' and node y', and conn(x',y') is the connectivity score between node x' and node y'; w1 is the weight coefficient, and its value is in the range of (0, 1); Where f(x') is the sample vector corresponding to node x', f(y') is the sample vector corresponding to node y', and @ is the vector dot multiplication operation; If there is a shortest path sp(x',y') between node x' and node y' in the undirected weighted graph G; then If node x' and node y' are directly connected in the undirected weighted graph G, then conn(x',y') = 1; Find all nodes whose comprehensive local density rho(x') is larger than that of x', and form a set, which is recorded as the large density set Nh(x'). For each node y'' in Nh(x'), calculate the distance d(x',y'') between x' and y'', and take the minimum value of the distance between all nodes in Nh(x') and x', de(x') = min(d(x',y'')); For each node x', calculate a clustering factor ga(x') = rho(x') × de(x'), sort all nodes in descending order according to the value of the clustering factor, start from the node with the largest clustering factor value, divide it and its nearest neighbors into a cluster, then move to the next undivided node with the largest clustering factor value, and repeat until all nodes are divided into clusters; For each cluster, it is regarded as a simple geometry; for each node belonging to the cluster, its membership is defined as the inverse of the distance between the corresponding node and the central node of the corresponding cluster; the combination of the membership of all simple geometries and their corresponding nodes constitutes a fuzzy geometric graph Gg.

6. The intelligent time-delay closing electric energy meter control platform according to claim 5 is characterized in that: The method of performing the optimization and dimensionality reduction process includes: Define a low-dimensional space and determine the target dimension d of the low-dimensional space; for each sample x in the data set X, take it as a data point and project it into the d-dimensional space to obtain a d-dimensional vector as the initial position of the data point in the low-dimensional space; in the low-dimensional space, based on the initial position of the data point, construct a low-dimensional fuzzy simple geometric graph G_low; Define the loss function F; F=-∑ (x∈X) ∑ (s∈Gg) [mh(x,s)×log(ml(x,s))]+λ×∑ (x∈X) [KL(Ph(x)||Pl(x))]; where mh(x,s) is the membership of x in the fuzzy geometric graph Gg to the convex hull shape s, and ml(x,s) is the membership of x in the low-dimensional fuzzy simple geometric graph G_low to the convex hull shape s; λ is the weight coefficient of the regularization term, Ph(x) is the neighbor distribution of x in the fuzzy geometric graph Gg, and Pl(x) is the neighbor distribution of x in the low-dimensional fuzzy simple geometric graph G_low; KL() represents the KL divergence; based on the loss function, the positions of data points in the low-dimensional space are iteratively optimized to minimize the loss function. In each iterative optimization, based on the current position of the data point, a low-dimensional fuzzy simple geometric graph is constructed, the value of the current loss function is calculated, and the position of each data point is back-propagated to obtain the gradient, and the optimizer is used to update the position of each data point according to the gradient; repeat until the preset maximum number of iterations is reached; the final low-dimensional fuzzy simple geometric graph is the multidimensional data space.

7. The intelligent time-delay closing electric energy meter control platform according to claim 6 is characterized in that: The evolution rule is: if a grid point in the live state has 2 to 3 neighbors in the live state, it will remain in the live state in the next generation; if a grid point in the live state has less than 2 or more than 3 neighbors in the live state, it will become dead in the next generation; if a grid point in the dead state has exactly 3 neighbors in the live state, it will become alive in the next generation; Define the length of the time step. In each iteration, for each grid point, find its neighbors in the grid and calculate their states. According to the evolution rules, determine the state of the corresponding grid point in the next time step, and update the states of all grid points at the same time. And at each time step, record the number of grid points whose states are alive. The grid whose states are iteratively updated is the two-dimensional grid evolution model.

8. The intelligent time-delay closing electric energy meter control platform according to claim 7 is characterized in that: The calculation formula for activity is: Where T is the number of time steps in the time window, t is the index of the time step, S(u,t) is the state of grid point u at time step t, Nac(u,t) is the number of active neighbors around grid point u at time step t, α is the neighborhood influence coefficient, and its value is in the range of (0, 1); A(u) is the activity of grid point u; is the step size control parameter; Convert the activity matrix into a binary image, and set the pixels corresponding to the potential peak area on the binary image to 1, and the others to 0; obtain a preliminary binary image; apply dilation and erosion operations to the preliminary binary image; obtain the potential peak area in the middle section; Define an overlap threshold; if the overlap between the current time window and a previous time window is higher than the overlap threshold, then record the duration between the two time windows; A minimum duration threshold mdu is set, and those mid-segment potential peak areas whose duration exceeds mdu are marked as stable areas. The tracking period and average threshold are set. For each pair of stable areas, their average overlap during the entire tracking period is calculated. If the average overlap is higher than the average threshold, the corresponding two stable areas are merged, that is, the union of the two corresponding stable areas in all time windows is calculated, and this union is used as the new merged area. The new merged area is morphologically processed to obtain the final power consumption peak area.

9. The intelligent time-delay closing electric energy meter control platform according to claim 8 is characterized in that: The weight coefficient of the conversion rate; is the historical impact weight coefficient; ΔP(t') is the power change at time t', Δt is the time interval; ti is the historical delay time point; τ is the time attenuation constant; in, is the importance weight coefficient of the f-th user, U_f(t') is the number of the f-th user at time t'; T_f(t') is the cumulative delay time of the f-th user at time t'; Wherein, θ is the load balance coefficient; Hmax is the maximum load value during the observation period, H′(t') is the actual load value at time t' after the delayed closing is implemented, and Hmin is the minimum load value during the observation period; The formula for adjusting the radius of the sphere is: Among them, R_new is the adjusted radius, R_p1 is the radius of sphere p1 corresponding to the optimal candidate solution, R_p2 is the sphere p2 with the largest function value of the candidate solution other than sphere p1 selected from all spheres; ε is the random perturbation factor, and its value range is (-0.1, 0.1); All spheres are sorted, spheres with smaller function values ​​are eliminated, and spheres with adjusted radius are used as new spheres; the change rate of the optimal candidate solution is iteratively and repeatedly calculated. If the change rate is less than the preset change threshold for K consecutive times, the last optimal candidate solution is used as the final solution, and the decision variables corresponding to the final solution are extracted to obtain the optimal delayed closing time and duration.

10. A method for controlling an intelligent time-delay closing electric energy meter, which is implemented based on the intelligent time-delay closing electric energy meter control platform according to any one of claims 1 to 9, characterized in that: include: Step 1: Obtain electricity consumption data and electricity consumption environment data of the electric energy meter; Step 2: construct a multidimensional data space based on the electricity consumption data and electricity consumption environment data of the energy meter; Step 3: Based on the multidimensional data space, a two-dimensional grid evolution model is constructed; Step 4, identifying the peak power consumption area in the two-dimensional grid evolution model; Step 5: For the identified peak power consumption area, the improved sphere optimization algorithm is used to determine the optimal delayed closing time and duration.

Images