An analysis and decision-making system and method based on power grid monitoring data
Through an analysis and decision-making system based on grid monitoring data, the future electricity consumption of the power consumption area is predicted and power allocation instructions are generated, and the blind spots of power supply gap evaluation in power dispatch are solved, ensuring the power supply stability of the power consumption area and the efficient utilization of power grid resources. .
Patent Information
- Application Number
- CN202410191313.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-02-21
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2044-02-21
AI Technical Summary
The current power dispatching operation relies too much on empirical judgment, which makes it difficult to timely and accurately evaluate the power supply gap in emergencies, resulting in unstable power supply in the power consumption area, affecting the quality of electricity consumption and reducing the efficiency of grid resource utilization.
Based on the analysis and decision-making system of power grid monitoring data, it generates power allocation instructions by predicting the future electricity consumption of the power consumption area, and uses the transmission grid topology diagram, node evaluation coefficient and transmission line evaluation coefficient to determine the optimal power station and transmission line to ensure that Meet electricity demand.
Identify in advance whether the power station meets the electricity demand, select a suitable backup power station to assist in power supply, improve the power consumption stability of the power consumption area, and ensure the efficient utilization of power grid resources by screening the most stable and minimal loss transmission lines.
Smart Images

Figure CN118054560B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of power grid monitoring, and in particular to an analysis and decision-making system and method based on power grid monitoring data. Background Art
[0002] As the scale of the power system continues to expand, the grid structure is becoming increasingly complex; however, current power dispatching and operational decisions are too dependent on the experience and instinctive judgment of dispatchers, resulting in certain blind spots and delays.
[0003] In particular, when faced with emergencies such as power supply gaps, if it is impossible to accurately assess the situation and make dispatch decisions in a timely manner; due to the lack of objective basis, the power dispatching department finds it difficult to understand the scale of the power supply gap in a timely manner, and the dispatching efficiency is low.
[0004] When selecting backup transmission lines to supplement power supply, lines with large losses or insufficient transmission capacity may be randomly selected, which may lead to unstable power supply or shortage of supply in power-consuming areas, seriously affecting the quality of electricity use.
[0005] If things continue like this, it will also reduce the efficiency of grid resource utilization and increase economic costs. Summary of the invention
[0006] The purpose of the present invention is to provide an analysis and decision-making system and method based on power grid monitoring data to solve the shortcomings of the background technology.
[0007] In order to achieve the above object, the present invention provides the following technical solution: an analysis and decision-making method based on power grid monitoring data, applied to a decision-making server, comprising:
[0008] According to the power consumption data of the w-th power consumption area, the power consumption of the w-th power consumption area in the future time T is predicted; the sum of the power consumption of the w power consumption areas is used as the expected power generation of the i-th power station in the future time T; the power consumption data is the power consumption data corresponding to the w power consumption areas collected by the power generation server, w is an integer greater than or equal to 1, there are i power stations, and i is an integer greater than 1;
[0009] The power generation data includes the current power generation and remaining power generation of the ith power station. The sum of the current power generation and the remaining power generation is marked as the total power generation. The expected power generation is compared and analyzed with the total power generation to determine whether to generate a power dispatch instruction. The power generation data is the power generation data of the power stations corresponding to the w power consumption areas collected by the power generation server.
[0010] If a power dispatch instruction is generated, w power consumption areas are taken as the starting point, and the power station other than the i-th substation is taken as the pending terminal power station, and the absolute value of the difference between the total power generation and the expected power generation is taken as the supplementary power; according to the transmission network topology diagram, r node evaluation coefficients, c transmission line evaluation coefficients, c remaining power and supplementary power, the final power station and the final transmission line are determined from the pending terminal power stations, and the supplementary power is taken as the additional power generation of the final power station at the future time T, and the power is transmitted to the w-th power consumption area through the final transmission line;
[0011] The node evaluation coefficient is the transmission data of r transmission network nodes collected by the transmission network server, and the node evaluation coefficients of the corresponding transmission network nodes are generated one by one according to the r transmission data; the transmission line evaluation coefficient is the input wire data of c input wires connected to the input ends of r transmission network nodes collected by the transmission network server, and the transmission line evaluation coefficients of the corresponding input wires are generated one by one according to the c input wire data, r and c are both integers greater than 1; the input wire data includes residual power.
[0012] Furthermore, the method for determining whether to generate a power allocation instruction includes:
[0013] If the difference between the estimated power generation and the total power generation is less than 0, a power dispatch instruction is generated; if the difference between the estimated power generation and the total power generation is greater than or equal to 0, no power dispatch instruction is generated.
[0014] Furthermore, the transmission data include transformer performance coefficient, load remaining capacity and power outage rate;
[0015] The node evaluation coefficient generation method includes:
[0016]
[0017] In the formula, JD r is the node evaluation coefficient of the rth transmission network node; SR r is the load remaining capacity of the rth transmission network node; XL r is the transformer performance factor of the rth transmission network node; WD r is the power supply interruption rate of the rth transmission network node, power supply interruption rate = (power outage time / total time) × 100%; All are preset weights, and their values are all greater than 0.
[0018] Furthermore, the transformer performance coefficient generation method includes:
[0019]
[0020] Where F is the total number of transformers in the rth transmission network node, YO rfis the coolant height of the fth transformer in the rth transmission network node; WZ rf is the fault-free interval time of the fth transformer in the rth transmission network node; CP rf is the CPU real-time load rate of the f-th transformer in the r-th transmission network node; η 1 , η 2 With η 3 All are preset weights;
[0021] The method for calculating the remaining load capacity includes:
[0022]
[0023] In the formula, SJ rf is the design capacity of the fth transformer in the rth transmission network node, design capacity = rated capacity × normal operating load rate; SF rf is the actual load of the fth transformer in the rth transmission network node.
[0024] Furthermore, the input power line data also includes power transmission loss; the transmission line evaluation coefficient calculation method includes:
[0025]
[0026] Where, SP q is the transmission line evaluation coefficient of the qth input line; SH q is the remaining power of the qth input line; SG q is the transmission loss of the qth input line; 1 With λ 2 The preset weights.
[0027] Furthermore, the calculation method of the residual power of the input line includes:
[0028] Query the design capacity of the input wire;
[0029] Get real-time load current;
[0030] Calculate the real-time load utilization, that is, the percentage of current load current to design capacity;
[0031] Calculate the remaining capacity: design capacity - real-time load current = remaining capacity;
[0032] Convert the remaining capacity into remaining power, obtain the voltage U of the input wire, and the available current I is the remaining capacity; the remaining power P of the input line = U × (I-Ie), where Ie is the reserved current.
[0033] Furthermore, the method for generating the transmission loss includes:
[0034] Input the transmission line loss characteristic data collected in real time into the pre-trained loss calculation model to obtain the transmission loss; wherein the transmission line loss characteristic data includes the transmission line cross-sectional area, the transmission line length, the ambient temperature and the transmission line type; the transmission line type includes gas termination and ground termination;
[0035] The training methods for the loss calculation model include:
[0036] Collect V sets of training data, where V is an integer greater than 1, and the training data includes transmission line loss characteristic data and transmission loss corresponding to the transmission line loss characteristic data.
[0037] The characteristic vector of each group of transmission line loss characteristic data is used as the input of the loss calculation model. The loss calculation model uses the transmission loss corresponding to each group of transmission line loss characteristic data as the output, and uses the transmission loss actually corresponding to each group of transmission line loss characteristic data as the prediction target; minimizing the sum of the prediction errors of the transmission losses corresponding to all transmission line loss characteristic data is used as the training target; the loss calculation model is trained until the sum of the prediction errors reaches convergence, and the training is stopped.
[0038] Furthermore, based on the power consumption data of the wth power consumption area, a method for predicting the power consumption of the wth power consumption area at a future time T includes:
[0039] Input the electricity consumption data of the w-th electricity consumption area into the pre-built electricity consumption prediction model corresponding to the w-th electricity consumption area, and output the electricity consumption of the w-th electricity consumption area at a future time T. The electricity consumption prediction model is an RNN or LSTM neural network model;
[0040] Methods for building electricity consumption forecasting models include:
[0041] Collect time series electricity consumption data sets, which include m groups of historical electricity consumption data and historical electricity consumption, where m is an integer greater than 1. Historical electricity consumption data include timestamp, temperature characteristics, season characteristics, characteristics of the same period last week, weather characteristics, and weekend and workday characteristics; timestamp is the time when each electricity consumption data is recorded; temperature characteristics include the highest and lowest temperatures of the day; season characteristics include spring, summer, autumn, and winter; characteristics of the same period last week include electricity consumption on the same date and during the same period last week; weather characteristics include weather indicators of whether it rains; weekend and workday characteristics include binary features that distinguish weekend and workday electricity consumption patterns;
[0042] Preset prediction time step T, sliding step P 2 And the sliding window length P 1 ; Use the sliding window method to convert m groups of historical electricity consumption data into multiple training samples, one training sample corresponds to one label, and constitutes a set of training data;
[0043] The training data is used as the input of the electricity consumption prediction model, and the future electricity consumption after the predicted time step T is used as the output. The actual electricity consumption in the future after each time step T is used as the prediction target, and the prediction accuracy is used as the training target. The electricity consumption prediction model is trained to obtain an electricity consumption prediction model that meets the prediction accuracy.
[0044] Furthermore, the pending terminal power station is a power station whose difference between the total power generation at the future time T and the expected power generation is greater than the supplementary power; the method for determining the final power station and the final transmission line from the pending terminal power stations includes:
[0045] Step 1: According to the power grid topology, several transmission lines between the w-th power consumption area and the to-be-determined terminal power station are determined. The transmission lines include z input wires and x transmission network nodes, z∈c, x∈r;
[0046] Step 2: Divide the replenished power by the unit time to obtain the electric power, and remove the input wires whose residual power is less than or equal to the electric power;
[0047] Step 3: Screen the complete transmission lines. The complete transmission lines are b, where b is an integer greater than or equal to 1. The complete transmission lines are the power stations with the pending end points that can connect the w-th power consumption area through the transmission lines.
[0048] Step 4: If b is equal to 1, the complete transmission line obtained by screening in step 3 is used as the final transmission line, and the power station at the pending terminal of the final transmission line is used as the final power station;
[0049] If b is greater than 1, then in the complete transmission line, a complete transmission line whose average node evaluation coefficient is greater than a preset node evaluation coefficient average threshold and whose standard deviation of the node evaluation coefficient is less than a preset node evaluation coefficient standard deviation threshold is selected and determined as the first target transmission line, and the power station connected to the first target transmission line is taken as the first target power station;
[0050] Step 5: There are s first target transmission lines, s∈b. If s is 1, the first target power station is used as the final power station, and the first target transmission line is used as the final transmission line;
[0051] If s is greater than 1, the first target transmission line whose average transmission line evaluation coefficient is greater than the preset transmission line evaluation coefficient average threshold and whose standard deviation of the transmission line evaluation coefficient is the smallest is selected as the second target transmission line, the power station connected to the second target transmission line is selected as the second target power station, the second target power station is selected as the final power station, and the second target transmission line is selected as the final transmission line.
[0052] An analysis and decision-making system based on power grid monitoring data, applied to a decision-making server, includes:
[0053] The power consumption prediction module is used to predict the power consumption of the w-th power consumption area at a future time T according to the power consumption data of the w-th power consumption area; the sum of the power consumption of the w power consumption areas is used as the estimated power generation of the i-th power station at the future time T; the power consumption data is the power consumption data corresponding to the w power consumption areas collected by the power generation server, w is an integer greater than or equal to 1, there are i power stations, and i is an integer greater than 1;
[0054] Instruction generation module, power generation data includes the current power generation and remaining power generation of the ith power station, and the sum of the current power generation and the remaining power generation is marked as the total power generation; the expected power generation is compared and analyzed with the total power generation to determine whether to generate a power dispatch instruction; the power generation data is the power generation data of the power stations corresponding to the w power consumption areas collected by the power generation server;
[0055] The grid dispatching module, if a power dispatching instruction is generated, takes w power consumption areas as the starting point, and the power station other than the i-th substation as the pending terminal power station, and takes the absolute value of the difference between the total power generation and the expected power generation as the supplementary power; according to the transmission network topology, r node evaluation coefficients, c transmission line evaluation coefficients, c remaining power and supplementary power, the final power station and the final transmission line are determined from the pending terminal power stations, and the supplementary power is taken as the additional power generation of the final power station at the future time T, and is transmitted to the w-th power consumption area through the final transmission line;
[0056] The node evaluation coefficient is the transmission data of r transmission network nodes collected by the transmission network server, and the node evaluation coefficients of the corresponding transmission network nodes are generated one by one according to the r transmission data; the transmission line evaluation coefficient is the input wire data of c input wires connected to the input ends of r transmission network nodes collected by the transmission network server, and the transmission line evaluation coefficients of the corresponding input wires are generated one by one according to the c input wire data, r and c are both integers greater than 1; the input wire data includes residual power.
[0057] In the above technical solution, the technical effects and advantages provided by the present invention are:
[0058] Through the time series electricity consumption data set, the electricity consumption in the future can be predicted, and it can be known in advance whether the corresponding power station can meet the electricity demand in the future. If it cannot meet the electricity demand, other power stations can be selected in advance to assist in power supply and ensure the electricity demand of the power consumption area;
[0059] Secondly, among the multiple transmission lines between other power stations and power consumption areas, r node evaluation coefficients, c transmission line evaluation coefficients and c residual powers are used to screen out the transmission lines with the most stable transmission performance, the smallest loss and the ability to meet the power consumption needs of the power consumption area, thereby improving the stability of power consumption in the power consumption area. BRIEF DESCRIPTION OF THE DRAWINGS
[0060] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.
[0061] Figure 1 The figure is a schematic diagram of an analysis and decision-making system based on power grid monitoring data of the present invention.
[0062] Figure 2 For the present invention Figure 1 Flowchart of an analysis and decision-making method based on power grid monitoring data. DETAILED DESCRIPTION
[0063] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are 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.
[0064] It should be noted that when an element is referred to as being "fixed to" another element, it may be directly on the other element or there may be an intermediate element. When an element is considered to be "connected to" another element, it may be directly connected to the other element or there may be an intermediate element at the same time.
[0065] Example 1
[0066] See also Figure 1 As shown, this embodiment is an analysis and decision-making system based on power grid monitoring data, including a decision-making server, a power generation server and a transmission network server, and the decision-making server, the power generation server and the transmission network server are connected by wire and / or wireless;
[0067] The power generation server is used to collect the power generation data of the power station and the power consumption data of w power consumption areas corresponding to the power station, where w is a positive integer greater than or equal to 1. The specific value of w is determined according to the preset relationship between the power station and the power consumption area. For example, the power generated by a power station supplies the power consumption of three power consumption areas, and the number of power consumption areas corresponding to the power station is 3, that is, w is 3, and the number of power generation servers is i. The power generation data of the corresponding power stations are collected one by one, that is, one power station is equipped with a power generation server, and the collected power generation data and power consumption data are sent to the decision server;
[0068] The electricity consumption area is the administrative division area, such as province, city, county, town, etc.;
[0069] The transmission network server is used to collect transmission data of r transmission network nodes and input wire data of c input wires connected to the input terminals of the r transmission network nodes, where the input wires are connected to the current input terminals of the transmission network nodes; generate node evaluation coefficients of corresponding transmission network nodes one by one according to the r transmission data, generate transmission line evaluation coefficients of corresponding input wires one by one according to the input wire data of the c input wires, and send the transmission line evaluation coefficients and the rth node evaluation coefficient to the decision server, where the transmission network node is a substation in the transmission network, and r is an integer greater than 1;
[0070] Transmission data include transformer performance coefficient, load remaining capacity and power outage rate; input power line data include remaining power and transmission loss;
[0071] The node evaluation coefficient generation method includes:
[0072]
[0073] In the formula, JD r is the node evaluation coefficient of the rth transmission network node; SR r is the load remaining capacity of the rth transmission network node, SR r The larger the value, the more stable the performance of the corresponding transmission network node when the transmission task is temporarily increased, and the greater the fault tolerance space; XL r is the transformer performance factor of the rth transmission network node, XL r The larger the WD is, the lower the transmission and distribution loss of the corresponding transmission network node is, the more stable the output voltage is, and the power efficiency is improved; r is the power supply interruption rate of the rth transmission network node, power supply interruption rate = (power outage time / total time) × 100%; the smaller the power supply interruption rate, the greater the proportion of normal power supply time of the transmission network node within a certain period of time, and the better the power supply reliability; All are preset weights, and their values are all greater than 0; the preset weight values are set by technicians in this field according to actual conditions; JD rThe larger the value, the more stable the power transmission performance of the corresponding transmission network node.
[0074] Transformer performance coefficient generation methods include:
[0075]
[0076] In the formula, F is the total number of transformers in the rth transmission network node, YO rf is the coolant height of the fth transformer in the rth transmission network node; WZ rf is the fault-free interval time of the fth transformer in the rth transmission network node; CP rf is the CPU real-time load rate of the f-th transformer in the r-th transmission network node; η 1 , η 2 With η 3 All of them are preset weights, and the values are selected and set by technicians in this field according to actual conditions.
[0077] The calculation methods of load remaining capacity include:
[0078]
[0079] In the formula, SJ rf is the design capacity of the fth transformer in the rth transmission network node, design capacity = rated capacity × normal operating load rate. For example, the rated capacity of the transformer is 1000KVA, and the normal operating load rate is 75%, then the design capacity is: 1000×75%=750KVA; SF rf is the actual load of the fth transformer in the rth transmission network node.
[0080] The calculation method of transmission line evaluation coefficient includes:
[0081]
[0082] In the formula, SP q is the transmission line evaluation coefficient of the qth input line; SH q is the residual power of the qth input line. The larger the residual power, the better the performance of buffering load changes and absorbing load fluctuations. When the transmission task is temporarily increased, there is enough capacity to transmit electricity. SG q is the transmission loss of the qth input line. The smaller the transmission loss, the higher the energy utilization rate, the smaller the voltage drop, the more stable the power supply, and the lower the transmission cost; 1 With λ 2 The weights are preset, and the values are selected by technicians in this field according to actual conditions.
[0083] The calculation method of the residual power of the input line includes:
[0084] Query the design capacity of the input wire, that is, the maximum current capacity that the input wire can theoretically transmit;
[0085] Obtain real-time load current data by using the transmission grid system to monitor the real-time load current passing through the input wires;
[0086] Calculate the real-time load utilization, that is, the percentage of current load current to design capacity, that is, real-time load current / design capacity;
[0087] Calculate the remaining capacity: design capacity - real-time load current = remaining capacity;
[0088] Convert the remaining capacity into remaining power, obtain the voltage U of the input wire, and the available current I is the remaining capacity; the remaining power P of the input line = U × (I-Ie), Ie is the reserved current used to prevent sudden load increase.
[0089] Transmission losses are generated by:
[0090] The transmission line loss characteristic data collected in real time is input into the pre-trained loss calculation model to obtain the transmission loss; wherein, the transmission line loss characteristic data includes the cross-sectional area of the transmission line, the length of the transmission line, the ambient temperature and the type of the transmission line; the increase in ambient temperature will increase the resistance of the conductor, thereby affecting the line loss; the types of transmission lines include gas-terminated and ground-terminated lines. The harmful electromagnetic field intensity of the gas-terminated line is lower, the line current distribution is more uniform, the line body temperature rise is also more uniform, and the line loss is relatively low; the ground-terminated line has a large grounding resistance, which introduces external electromagnetic interference, the harmful electromagnetic field intensity of the line is higher, and the line loss is relatively high;
[0091] The training methods for the loss calculation model include:
[0092] Collect V sets of training data, where V is an integer greater than 1, and the training data includes transmission line loss characteristic data and transmission loss corresponding to the transmission line loss characteristic data.
[0093] The characteristic vector of each group of transmission line loss characteristic data is used as the input of the loss calculation model. The loss calculation model uses the transmission loss corresponding to each group of transmission line loss characteristic data as the output, and uses the transmission loss actually corresponding to each group of transmission line loss characteristic data as the prediction target; minimizing the sum of the prediction errors of the transmission losses corresponding to all transmission line loss characteristic data is used as the training target; the loss calculation model is trained until the sum of the prediction errors reaches convergence and the training is stopped. The loss calculation model is a regression prediction model or other suitable machine learning model, which is not specifically limited here.
[0094] Among them, the calculation formula of the prediction error is ZW V =(α V-μ V ) 2 , where ZW V is the prediction error, V is the group number of the characteristic vector of the transmission line loss characteristic data, α V is the actual transmission loss corresponding to the Vth group of transmission line loss characteristic data, μ V Predict the corresponding transmission loss for the Vth group of transmission line loss characteristic data.
[0095] The decision server includes:
[0096] The power consumption prediction module is used to predict the power consumption of the wth power consumption area at a future time T based on the power consumption data of the wth power consumption area; the sum of the power consumption of the w power consumption areas is used as the estimated power generation of the i-th power station at a future time T;
[0097] Instruction generation module, the power generation data includes the current power generation and the remaining power generation of the ith power station, and the sum of the current power generation and the remaining power generation is marked as the total power generation; the expected power generation is compared and analyzed with the total power generation to determine whether to generate a power dispatch instruction;
[0098] If the difference between the estimated power generation and the total power generation is less than 0, a power dispatch instruction is generated; if the difference between the estimated power generation and the total power generation is greater than or equal to 0, no power dispatch instruction is generated; this indicates that the total power generation of the i-th power station can meet the power consumption of w power consumption areas at the future time T;
[0099] The grid dispatching module, if generating a power dispatching instruction, takes w power consumption areas as the starting point, and the power station other than the i-th substation as the pending terminal power station, and takes the absolute value of the difference between the total power generation and the expected power generation as the supplementary power;
[0100] According to the transmission network topology, node evaluation coefficient, transmission line evaluation coefficient, remaining power and supplementary power, the final power station and the final transmission line are determined from the pending terminal power stations, and the supplementary power is used as the additional power generation of the final power station at the future time T, and the power is transmitted to the w-th power consumption area through the final transmission line;
[0101] According to the power consumption data of the w-th power consumption area, a method for predicting the power consumption of the w-th power consumption area at a future time T includes:
[0102] Input the electricity consumption data of the w-th electricity consumption area into the pre-built electricity consumption prediction model corresponding to the w-th electricity consumption area, and output the electricity consumption of the w-th electricity consumption area at a future time T. The electricity consumption prediction model is an RNN / LSTM neural network model;
[0103] Methods for building electricity consumption forecasting models include:
[0104] Collect time series electricity consumption data sets, which include m groups of historical electricity consumption data and historical electricity consumption, where m is an integer greater than 1. Historical electricity consumption data include timestamp, temperature characteristics, seasonal characteristics, characteristics of the same period last week, weather characteristics, and weekend and working day characteristics; timestamp is the time when each electricity consumption data is recorded, such as time dimensions such as day, hour, and minute; temperature characteristics include the highest and lowest temperatures of the day; seasonal characteristics include spring, summer, autumn, and winter; characteristics of the same period last week include electricity consumption on the same date and during the same period last week, reflecting the periodicity of electricity consumption patterns; weather characteristics include weather indicators of whether it rains; weekend and working day characteristics include binary features that distinguish weekend and working day electricity consumption patterns; comprehensively considering the time series electricity consumption data that affects future electricity consumption, it can be used to estimate the power supply gap in the future, provide objective basis for the power dispatching department, timely and accurately assess the scale of the power supply gap, and the system automatically makes dispatching decisions, reducing dependence on the experience of dispatching personnel, with a high degree of intelligence and strong timeliness, avoiding misjudgments caused by human operations;
[0105] Convert the historical electricity consumption data into feature vectors as follows:
[0106] X 1 =[T 1 ,WZM 1 ,WZN 1 ,JZ 1 ,ZS 1 ,TZ 1 ,GZ 1 ], X 1 The corresponding power consumption is Y 1 ;
[0107] X 1 is the first set of eigenvectors; T 1 The timestamp in the first set of feature vectors, such as hours or minutes; WZM 1 is the maximum temperature of the day in the first set of eigenvectors; WZN 1 is the lowest temperature of the day in the first set of eigenvectors; JZ 1 is the season in the first set of eigenvectors, such as spring is 1, summer is 2, autumn is 3, and winter is 4; ZS 1 TZ is the electricity consumption at the same time last week in the first set of feature vectors; 1 is the weather in the first set of eigenvectors, such as 1 if it rains and 0 if it does not rain; GZ 1 In the first set of feature vectors, weekends or weekdays are 1 and weekdays are 0;
[0108] Preset prediction time step T, sliding step P 2 And the sliding window length P 1; Use the sliding window method to convert m groups of historical electricity consumption data into multiple training samples, one training sample corresponds to one label, and constitutes a set of training data, P2 and P1 are both integers greater than 1;
[0109] The training data is used as the input of the electricity consumption prediction model, and the future electricity consumption after the predicted time step T is used as the output. The actual future electricity consumption after each time step T is used as the prediction target, and the prediction accuracy is used as the training target. The electricity consumption prediction model is trained to obtain an electricity consumption prediction model that meets the prediction accuracy. That is, the electricity consumption prediction model is built;
[0110] The pending terminal power station is the power station whose total power generation at the future time T has a difference with the expected power generation that is greater than the supplementary power;
[0111] The pending terminal power station is a power station whose total power generation at a future time T differs from the expected power generation by a larger amount than the supplementary power. The method for determining the final power station and the final transmission line from the pending terminal power stations includes:
[0112] Step 1: According to the power grid topology, several transmission lines between the w-th power consumption area and the undetermined terminal power station are determined. The transmission lines include z input wires and x transmission network nodes, z∈c, x∈r;
[0113] Step 2: The supplementary power is divided by the unit time to obtain the electric power, and the input wires whose residual power is less than or equal to the electric power are eliminated, so that the eliminated input wires have sufficient capacity to increase the required transmission power in the future time T; before comparing the electric power with the residual power, the electric power and the residual power units need to be converted to be consistent; here the inventors take into account that the performance of the input wires in the actual scenario is different, that is, there are lines that do not match the demand for electricity per unit time in the power consumption area. According to the demand for electricity per unit time in the power consumption area, the transmission lines that do not match the demand for electricity per unit time in the power consumption area are eliminated to ensure the quality of power supply to the power consumption area in the later stage;
[0114] Step 3: Screen the complete transmission lines. The complete transmission lines are b, where b is an integer greater than or equal to 1. The complete transmission lines are the transmission lines that can connect the power station at the pending terminal to the w-th power consumption area, ensuring that the transmission lines between the power station at the pending terminal and the w-th power consumption area are unobstructed.
[0115] Step 4: If b is equal to 1, the complete transmission line obtained by screening in step 3 is used as the final transmission line, and the power station at the pending terminal of the final transmission line is used as the final power station;
[0116] If b is greater than 1, then in the complete transmission line, a complete transmission line whose average node evaluation coefficient is greater than a preset node evaluation coefficient average threshold and whose standard deviation of the node evaluation coefficient is less than a preset node evaluation coefficient standard deviation threshold is selected and determined as the first target transmission line, and the power station connected to the first target transmission line is taken as the first target power station; the average node evaluation coefficient reflects the transmission performance of the node in the complete transmission line, and the standard deviation of the node evaluation coefficient reflects the degree of dispersion between the evaluation coefficient of each node in the complete transmission line and the average node evaluation coefficient threshold. The smaller the standard deviation of the node evaluation coefficient is, the more stable the transmission performance of each node is;
[0117] Step 5: There are s first target transmission lines, s∈b. If s is 1, the first target power station is used as the final power station, and the first target transmission line is used as the final transmission line;
[0118] If s is greater than 1, the first target transmission line whose average value of the transmission line evaluation coefficient is greater than the preset average threshold of the transmission line evaluation coefficient and whose standard deviation of the transmission line evaluation coefficient is the smallest is selected as the second target transmission line, the power station connected to the second target transmission line is taken as the second target power station, the second target power station is taken as the final power station, and the second target transmission line is taken as the final transmission line; the node evaluation coefficient average threshold, the node evaluation coefficient standard deviation threshold and the wire evaluation coefficient average threshold are determined by technical personnel according to specific application requirements.
[0119] In summary, the present embodiment is an analysis and decision-making system based on power grid monitoring data. Through the time series power consumption data set, it is used for the power consumption in the future time, and it is known in advance whether the corresponding power station can meet the power demand in the future time. If the power demand cannot be met, other power stations can be selected in advance to assist in power supply, so as to ensure the power demand of the power consumption area;
[0120] Secondly, when selecting backup transmission lines to supplement power supply, among the multiple transmission lines between other power stations and power consumption areas, r node evaluation coefficients, c transmission line evaluation coefficients and c residual powers are used to eliminate transmission lines with insufficient transmission capacity from multiple transmission lines, and then select the transmission lines with the most stable transmission performance, the smallest loss, and the ability to meet the power consumption needs of the power consumption area, so as to ensure the stability of power supply in the power consumption area and improve the quality of power consumption.
[0121] Example 2
[0122] See also Figure 2 As shown, the part not described in detail in this embodiment is described in Example 1, which provides an analysis and decision-making method based on power grid monitoring data, which is applied to a decision server and includes:
[0123] According to the power consumption data of the w-th power consumption area, the power consumption of the w-th power consumption area in the future time T is predicted; the sum of the power consumption of the w power consumption areas is used as the expected power generation of the i-th power station in the future time T; the power consumption data is the power consumption data corresponding to the w power consumption areas collected by the power generation server, w is an integer greater than or equal to 1, there are i power stations, and i is an integer greater than 1;
[0124] The power generation data includes the current power generation and remaining power generation of the ith power station. The sum of the current power generation and the remaining power generation is marked as the total power generation. The expected power generation is compared and analyzed with the total power generation to determine whether to generate a power dispatch instruction. The power generation data is the power generation data of the power stations corresponding to the w power consumption areas collected by the power generation server.
[0125] If a power dispatch instruction is generated, w power consumption areas are taken as the starting point, and the power station other than the i-th substation is taken as the pending terminal power station, and the absolute value of the difference between the total power generation and the expected power generation is taken as the supplementary power; according to the transmission network topology diagram, r node evaluation coefficients, c transmission line evaluation coefficients, c remaining power and supplementary power, the final power station and the final transmission line are determined from the pending terminal power stations, and the supplementary power is taken as the additional power generation of the final power station at the future time T, and the power is transmitted to the w-th power consumption area through the final transmission line;
[0126] The node evaluation coefficient is the transmission data of r transmission network nodes collected by the transmission network server, and the node evaluation coefficients of the corresponding transmission network nodes are generated one by one according to the r transmission data; the transmission line evaluation coefficient is the input wire data of c input wires connected to the input ends of r transmission network nodes collected by the transmission network server, and the transmission line evaluation coefficients of the corresponding input wires are generated one by one according to the c input wire data, r and c are both integers greater than 1; the input wire data includes residual power.
[0127] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. An analysis and decision-making method based on power grid monitoring data, applied to a decision server, characterized in that: include: Based on the electricity consumption data of the w-th electricity consumption area, predict the electricity consumption of the w-th electricity consumption area at a future time T; The sum of the electricity consumption of w electricity consumption areas is taken as the estimated power generation of the i-th power station at the future time T; the electricity consumption data is the electricity consumption data corresponding to the w electricity consumption areas collected by the power generation server, w is an integer greater than or equal to 1, there are i power stations, and i is an integer greater than 1; The power generation data includes the current power generation and the remaining power generation of the i-th power station, and the sum of the current power generation and the remaining power generation is marked as the total power generation; Compare and analyze the expected power generation with the total power generation to determine whether to generate a power dispatch instruction; the power generation data is the power generation data of the power plants corresponding to the w power consumption areas collected by the power generation server; If a power dispatch instruction is generated, w power consumption areas are taken as the starting point, and the power station other than the i-th substation is taken as the pending terminal power station, and the absolute value of the difference between the total power generation and the expected power generation is taken as the supplementary power; according to the transmission network topology diagram, r node evaluation coefficients, c transmission line evaluation coefficients, c remaining power and supplementary power, the final power station and the final transmission line are determined from the pending terminal power stations, and the supplementary power is taken as the additional power generation of the final power station at the future time T, and the power is transmitted to the w-th power consumption area through the final transmission line; The pending destination power station is a power station whose total power generation at the future time T and the expected power generation have a difference greater than the supplementary power; The method of determining the final power station and the final transmission line from the pending terminal power stations includes: Step 1: According to the power grid topology, several transmission lines between the w-th power consumption area and the undetermined terminal power station are determined. The transmission lines include z input wires and x transmission network nodes, z∈c, x∈r; Step 2: Divide the replenished power by the unit time to obtain the electric power, and remove the input wires whose residual power is less than or equal to the electric power; Step 3: Screen the complete transmission lines. The complete transmission lines are b, where b is an integer greater than or equal to 1. The complete transmission lines are the power stations with the pending end points that can connect the w-th power consumption area through the transmission lines. Step 4: If b is equal to 1, the complete transmission line obtained by screening in step 3 is used as the final transmission line, and the power station at the pending terminal of the final transmission line is used as the final power station; If b is greater than 1, then in the complete transmission line, a complete transmission line whose average node evaluation coefficient is greater than a preset node evaluation coefficient average threshold and whose standard deviation of the node evaluation coefficient is less than a preset node evaluation coefficient standard deviation threshold is selected and determined as the first target transmission line, and the power station connected to the first target transmission line is taken as the first target power station; Step 5: There are s first target transmission lines, s∈b. If s is 1, the first target power station is used as the final power station, and the first target transmission line is used as the final transmission line; If s is greater than 1, the first target transmission line whose average transmission line evaluation coefficient is greater than the preset transmission line evaluation coefficient average threshold and whose standard deviation of the transmission line evaluation coefficient is the smallest is selected as the second target transmission line, the power station connected to the second target transmission line is selected as the second target power station, the second target power station is selected as the final power station, and the second target transmission line is selected as the final transmission line; The node evaluation coefficient is the transmission data of r transmission network nodes collected by the transmission network server, and the node evaluation coefficients of the corresponding transmission network nodes are generated one by one according to the r transmission data; the transmission line evaluation coefficient is the input wire data of c input wires connected to the input ends of r transmission network nodes collected by the transmission network server, and the transmission line evaluation coefficients of the corresponding input wires are generated one by one according to the c input wire data, r and c are both integers greater than 1; the input wire data includes residual power.
2. The analysis and decision-making method based on power grid monitoring data according to claim 1 is characterized in that: The method for determining whether to generate a power allocation instruction includes: If the difference between the estimated power generation and the total power generation is less than 0, a power dispatch instruction is generated; if the difference between the estimated power generation and the total power generation is greater than or equal to 0, no power dispatch instruction is generated.
3. The analysis and decision-making method based on power grid monitoring data according to claim 1 is characterized in that: The power transmission data includes transformer performance coefficient, load remaining capacity and power supply interruption rate; The node evaluation coefficient generation method comprises: In the formula, JD r is the node evaluation coefficient of the rth transmission network node; SR r is the load remaining capacity of the rth transmission network node; XL r is the transformer performance factor of the rth transmission network node; WD r is the power supply interruption rate of the rth transmission network node, power supply interruption rate = (power outage time / total time) × 100%; All are preset weights, and their values are all greater than 0.
4. The analysis and decision-making method based on power grid monitoring data according to claim 3 is characterized in that: The transformer performance coefficient generation method comprises: Where F is the total number of transformers in the rth transmission network node, YO rf is the coolant height of the fth transformer in the rth transmission network node; WZ rf is the fault-free interval time of the fth transformer in the rth transmission network node; CP rf is the CPU real-time load rate of the f-th transformer in the r-th transmission network node; η1, η2 and η3 are all preset weights; The method for calculating the remaining load capacity includes: Where SJrf is the design capacity of the f-th transformer in the r-th transmission network node, and the design capacity = rated capacity × normal operating load rate; SF rf is the actual load of the fth transformer in the rth transmission network node.
5. The analysis and decision-making method based on power grid monitoring data according to claim 1 is characterized in that: The input power line data also includes power transmission loss; the power transmission line evaluation coefficient calculation method includes: Where, SP q is the transmission line evaluation coefficient of the qth input line; SH q is the remaining power of the qth input line; SG q is the transmission loss of the qth input line; λ1 and λ2 are preset weights.
6. The analysis and decision-making method based on power grid monitoring data according to claim 5 is characterized in that: The method for calculating the residual power of the input line includes: Query the design capacity of the input wire; Get real-time load current; Calculate the real-time load utilization, that is, the percentage of current load current to design capacity; Calculate the remaining capacity: design capacity - real-time load current = remaining capacity; Convert the remaining capacity into remaining power, obtain the voltage U of the input wire, and the available current I is the remaining capacity; the remaining power P of the input line = U × (I-Ie), where Ie is the reserved current.
7. The analysis and decision-making method based on power grid monitoring data according to claim 5 is characterized in that: The method for generating the transmission loss comprises: Input the transmission line loss characteristic data collected in real time into the pre-trained loss calculation model to obtain the transmission loss; wherein the transmission line loss characteristic data includes the transmission line cross-sectional area, the transmission line length, the ambient temperature and the transmission line type; the transmission line type includes gas termination and ground termination; The training method of the loss calculation model includes: Collect V sets of training data, where V is an integer greater than 1, and the training data includes transmission line loss characteristic data and transmission loss corresponding to the transmission line loss characteristic data. The characteristic vector of each group of transmission line loss characteristic data is used as the input of the loss calculation model. The loss calculation model uses the transmission loss corresponding to each group of transmission line loss characteristic data as the output, and uses the transmission loss actually corresponding to each group of transmission line loss characteristic data as the prediction target; minimizing the sum of the prediction errors of the transmission losses corresponding to all transmission line loss characteristic data is used as the training target; the loss calculation model is trained until the sum of the prediction errors reaches convergence, and the training is stopped.
8. The analysis and decision-making method based on power grid monitoring data according to claim 1 is characterized in that: According to the power consumption data of the wth power consumption area, the method for predicting the power consumption of the wth power consumption area at a future time T includes: Input the electricity consumption data of the w-th electricity consumption area into the pre-built electricity consumption prediction model corresponding to the w-th electricity consumption area, and output the electricity consumption of the w-th electricity consumption area at a future time T. The electricity consumption prediction model is an RNN or LSTM neural network model; Methods for building electricity consumption forecasting models include: Collect time series electricity consumption data sets, which include m groups of historical electricity consumption data and historical electricity consumption, where m is an integer greater than 1. Historical electricity consumption data include timestamp, temperature characteristics, season characteristics, characteristics of the same period last week, weather characteristics, and weekend and workday characteristics; timestamp is the time when each electricity consumption data is recorded; temperature characteristics include the highest and lowest temperatures of the day; season characteristics include spring, summer, autumn, and winter; characteristics of the same period last week include electricity consumption on the same date and during the same period last week; weather characteristics include weather indicators of whether it rains; weekend and workday characteristics include binary features that distinguish weekend and workday electricity consumption patterns; Preset the prediction time step T, sliding step P2 and sliding window length P1; convert m groups of historical electricity consumption data into multiple training samples using the sliding window method, where one training sample corresponds to one label and constitutes a group of training data; The training data is used as the input of the electricity consumption prediction model, and the future electricity consumption after the predicted time step T is used as the output. The actual electricity consumption in the future after each time step T is used as the prediction target, and the prediction accuracy is used as the training target. The electricity consumption prediction model is trained to obtain an electricity consumption prediction model that meets the prediction accuracy.
9. An analysis and decision-making system based on power grid monitoring data, applied to a decision-making server, characterized in that: include: The power consumption prediction module is used to predict the power consumption of the wth power consumption area at a future time T based on the power consumption data of the wth power consumption area; The sum of the electricity consumption of w electricity consumption areas is taken as the estimated power generation of the i-th power station at the future time T; the electricity consumption data is the electricity consumption data corresponding to the w electricity consumption areas collected by the power generation server, w is an integer greater than or equal to 1, there are i power stations, and i is an integer greater than 1; Instruction generation module, power generation data includes the current power generation and remaining power generation of the ith power station, and the sum of the current power generation and the remaining power generation is marked as the total power generation; the expected power generation is compared and analyzed with the total power generation to determine whether to generate a power dispatch instruction; the power generation data is the power generation data of the power stations corresponding to the w power consumption areas collected by the power generation server; The grid dispatching module, if a power dispatching instruction is generated, takes w power consumption areas as the starting point, and the power station other than the i-th substation as the pending terminal power station, and takes the absolute value of the difference between the total power generation and the expected power generation as the supplementary power; according to the transmission network topology, r node evaluation coefficients, c transmission line evaluation coefficients, c remaining power and supplementary power, the final power station and the final transmission line are determined from the pending terminal power stations, and the supplementary power is taken as the additional power generation of the final power station at the future time T, and is transmitted to the w-th power consumption area through the final transmission line; The pending destination power station is a power station whose total power generation at the future time T and the expected power generation have a difference greater than the supplementary power; The method of determining the final power station and the final transmission line from the pending terminal power stations includes: Step 1: According to the power grid topology, several transmission lines between the w-th power consumption area and the undetermined terminal power station are determined. The transmission lines include z input wires and x transmission network nodes, z∈c, x∈r; Step 2: Divide the replenished power by the unit time to obtain the electric power, and remove the input wires whose residual power is less than or equal to the electric power; Step 3: Screen the complete transmission lines. The complete transmission lines are b, where b is an integer greater than or equal to 1. The complete transmission lines are the power stations with the pending end points that can connect the w-th power consumption area through the transmission lines. Step 4: If b is equal to 1, the complete transmission line obtained by screening in step 3 is used as the final transmission line, and the power station at the pending terminal of the final transmission line is used as the final power station; If b is greater than 1, then in the complete transmission line, a complete transmission line whose average node evaluation coefficient is greater than a preset node evaluation coefficient average threshold and whose standard deviation of the node evaluation coefficient is less than a preset node evaluation coefficient standard deviation threshold is selected and determined as the first target transmission line, and the power station connected to the first target transmission line is taken as the first target power station; Step 5: There are s first target transmission lines, s∈b. If s is 1, the first target power station is used as the final power station, and the first target transmission line is used as the final transmission line; If s is greater than 1, the first target transmission line whose average transmission line evaluation coefficient is greater than the preset transmission line evaluation coefficient average threshold and whose standard deviation of the transmission line evaluation coefficient is the smallest is selected as the second target transmission line, the power station connected to the second target transmission line is selected as the second target power station, the second target power station is selected as the final power station, and the second target transmission line is selected as the final transmission line; The node evaluation coefficient is the transmission data of r transmission network nodes collected by the transmission network server, and the node evaluation coefficients of the corresponding transmission network nodes are generated one by one according to the r transmission data; the transmission line evaluation coefficient is the input wire data of c input wires connected to the input ends of r transmission network nodes collected by the transmission network server, and the transmission line evaluation coefficients of the corresponding input wires are generated one by one according to the c input wire data, r and c are both integers greater than 1; the input wire data includes residual power.
