CN118676940A - Flexible load aggregation characteristic identification and optimization method - Google Patents
Flexible load aggregation characteristic identification and optimization method Download PDFInfo
- Publication number
- CN118676940A CN118676940A CN202411139397.2A CN202411139397A CN118676940A CN 118676940 A CN118676940 A CN 118676940A CN 202411139397 A CN202411139397 A CN 202411139397A CN 118676940 A CN118676940 A CN 118676940A
- Authority
- CN
- China
- Prior art keywords
- time
- electricity price
- real
- response
- electricity
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for AC mains or AC distribution networks
- H02J3/12—Circuit arrangements for AC mains or AC distribution networks for adjusting voltage in AC networks by changing a characteristic of the network load
- H02J3/14—Circuit arrangements for AC mains or AC distribution networks for adjusting voltage in AC networks by changing a characteristic of the network load by switching loads on to, or off from, network, e.g. progressively balanced loading
- H02J3/144—Demand-response operation of the power transmission or distribution network
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for AC mains or AC distribution networks
- H02J3/003—Load forecast, e.g. methods or systems for forecasting future load demand
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J2310/00—The network for supplying or distributing electric power characterised by its spatial reach or by the load
- H02J2310/70—Load identification
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Economics (AREA)
- General Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Physics & Mathematics (AREA)
- Human Resources & Organizations (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- Health & Medical Sciences (AREA)
- Power Engineering (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Primary Health Care (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Life Sciences & Earth Sciences (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
本申请公开了一种柔性负荷聚合特性辨识与优化方法，涉及电力需求侧响应技术领域，包括：获取面向柔性负荷的电价量级表，电价量级表包括激励电价和对应的时间节点，选取初始激励电价进行首次输入到特性辨识模型中，依据初始激励电价按照量级规则依次选取剩余的激励电价进行输入，分别得到对应的实时响应用电量和实时负荷响应特征向量，将符合量级规则的激励电价对应的时间节点标记为时度节点；将不小于电量幅度阈值的实时响应用电量标记为电度节点；跨度节点是时度节点和电度节点相重合的节点；实现更精准的激励电价控制优化的效果；生成优化调控报告，提出针对性的优化调控策略，实现了最大化经济效益的效果。
The present application discloses a method for identifying and optimizing the characteristics of flexible load aggregation, which relates to the technical field of power demand side response, including: obtaining an electricity price magnitude table for flexible loads, the electricity price magnitude table including incentive electricity prices and corresponding time nodes, selecting an initial incentive electricity price for first input into a characteristic identification model, selecting the remaining incentive electricity prices in turn according to the magnitude rule based on the initial incentive electricity price for input, and obtaining corresponding real-time response electricity consumption and real-time load response characteristic vectors respectively, marking the time nodes corresponding to the incentive electricity prices that meet the magnitude rule as time nodes; marking the real-time response electricity consumption that is not less than the electricity amplitude threshold as electricity nodes; the span node is the node where the time node and the electricity node overlap; achieving a more accurate incentive electricity price control optimization effect; generating an optimization control report, proposing a targeted optimization control strategy, and achieving the effect of maximizing economic benefits.
Description
Technical Field
The invention relates to the technical field of power demand side response, in particular to a flexible load aggregation characteristic identification and optimization method.
Background
The development of demand side flexible resources which are not fully activated yet, the full grasp of the construction opportunity of the power spot market, the development of electricity price excitation type demand response technology and the development of the demand side flexible resources become important means for improving the demand side flexibility, and the demand side flexible resources mainly comprise a series of flexible loads, including electric automobiles, intelligent building structures, multi-energy micro-grids and the like. The resources generally have the characteristic of massive isomerism, the resource distribution is dispersed, and the large-scale controllable resources can be formed through efficient aggregation optimization treatment. In the prior art, a user is relied on to report operation parameters actively in modeling, so that modeling performance is affected by accuracy of the reported parameters, for example, when the operation parameters are distorted or malicious errors exist, a real system optimal scheme cannot be obtained whether a centralized direct load control algorithm or a distributed decomposition coordination algorithm is adopted.
The Chinese patent application No. 202211128637.X provides a method, a device and equipment for identifying and optimizing the aggregate characteristics of a non-invasive flexible load, which are used for acquiring a characteristic identification model and an elasticity estimation model for the flexible load, acquiring the excitation electricity price of the current round in real time, respectively inputting the excitation electricity price of the current round into the characteristic identification model and the elasticity estimation model to output real-time response electricity consumption and a real-time virtual elasticity matrix, finally acquiring the optimal excitation electricity price and the optimal response electricity consumption, realizing the aggregate optimization control of the non-invasive flexible load, and establishing the equivalent mapping relation of external characteristics by using a statistical method without depending on the active information reporting of a user.
However, in practical application, aiming at industries or industrial and commercial users with high power dependency, especially large-scale energy storage users, the demand level of electricity price is higher, the excitation electricity price cannot be accurately obtained when aggregation optimization is carried out in the prior art, and the maximized economic benefit is not realized on the premise of ensuring the load safety.
Disclosure of Invention
The application solves the problem that the excitation electricity price cannot be accurately regulated and controlled to reduce the economic benefit in the prior art by providing the flexible load aggregation characteristic identification and optimization method, and achieves the technical effects of accurately regulating and controlling the excitation electricity price and maximizing the economic benefit.
The application provides a flexible load aggregation characteristic identification and optimization method, which comprises the following steps:
S100: acquiring a characteristic identification model oriented to flexible load, wherein the input of the characteristic identification model is excitation electricity price, and the output of the characteristic identification model is real-time response electricity consumption and real-time load response characteristic vector; the real-time load response characteristic vector comprises a real-time response power change rate, load rate fluctuation, response sensitivity and response delay;
S200: acquiring a flexible load-oriented electricity price magnitude table, wherein the electricity price magnitude table comprises excitation electricity price and corresponding time nodes, selecting initial excitation electricity price to be input into a characteristic identification model for the first time, sequentially selecting residual excitation electricity price to be input according to the initial excitation electricity price and magnitude rule, respectively obtaining corresponding real-time response electricity consumption and real-time load response feature vectors, and marking the time nodes corresponding to the excitation electricity price conforming to the magnitude rule as time nodes;
S300: acquiring the output real-time response electricity consumption, presetting an electricity amplitude threshold according to the historical real-time response electricity consumption change rate, comparing the real-time response electricity consumption change rate with the electricity amplitude threshold, and marking a time node corresponding to the real-time response electricity consumption which is not smaller than the electricity amplitude threshold as an electricity node; the electric quantity amplitude threshold is set according to actual conditions and historical data and is used for measuring the change amplitude of the electric quantity change rate of the real-time response.
S400: generating a time curve according to different excitation electricity prices and time nodes, generating an electricity curve according to the real-time response electricity consumption and the electricity nodes, determining span nodes, acquiring the real-time load response feature vector of the span nodes, and generating an optimized regulation report; the span node is a node where a time node and an electricity node are overlapped;
S500: and carrying out aggregation optimization on the flexible load according to the optimization regulation report and updating an electricity price magnitude table.
Further, the magnitude rule is to select the excitation electricity price which is different in magnitude from the initial excitation electricity price or not smaller than a measurement threshold value as the excitation electricity price of the subsequent input by taking the initial excitation electricity price as a reference; and calculating the variation amplitude between the excitation electricity prices with the same magnitude, and comparing the variation amplitude with the measurement threshold value. The measurement threshold is set according to actual conditions and experience and is used for measuring the change amplitude of the excitation electricity price.
Further, two adjacent span nodes are obtained, a time area between the two span nodes is divided into a plurality of auxiliary control areas according to the excitation electricity price, and the same excitation electricity price is the same auxiliary control area; in each auxiliary control area, dividing the auxiliary control area into a plurality of segment areas according to the difference value of the exciting electricity price; the segment areas correspond to a newly divided excitation electricity price, and corresponding real-time electricity consumption and real-time load response characteristic vectors are obtained according to the new excitation electricity price; and generating an optimized regulation report according to the optimal excitation electricity price corresponding to the fragment area and the corresponding time range obtained by analyzing the real-time response electricity consumption and the real-time load response characteristic vector.
Further, the difference value is a difference value of excitation electricity prices among different auxiliary control areas, and is obtained by recording end electricity prices and initial electricity prices of the auxiliary control areas, and the auxiliary control areas with the difference value being more than 0, namely the end electricity prices being more than the initial electricity prices, are divided; and setting an auxiliary control area with the difference value smaller than 0, namely the ending electricity price smaller than the starting electricity price as a pending area.
Further, the segment areas are divided according to the next magnitude of the difference, the maximum number of the segment areas is not more than 10, new excitation electricity prices are set for each segment area after division is completed, the time length of the auxiliary control area is divided according to the number of the segment areas, and the initial time node and the end time node of each segment area are recorded.
Further, an initial time node of the segment area is obtained, a time range is gradually increased according to an extension step length, the real-time response electricity consumption change rate and the real-time load response characteristic vector are obtained according to the current time range and the new excitation electricity price, whether a preset condition is reached or not is judged, extension is stopped if the preset condition is reached or the maximum extension times are reached, and the time node at the moment is recorded as an end time node.
Further, the maximum extension frequency needs to be preset according to the time length and the extension step length of the auxiliary control area.
Further, the preset condition comprises a first threshold value and a second threshold value, and the first threshold value is set according to the real-time response power consumption change rate; the second threshold is set according to a comprehensive fluctuation value, the comprehensive fluctuation value is generated comprehensively according to the load rate fluctuation, the response sensitivity and the response delay, and if the real-time response electricity consumption change rate reaches the first threshold or the comprehensive fluctuation value reaches the second threshold, the extension is stopped.
Further, the time range of the segment area and the excitation electricity price are obtained, a preliminary reference value and an initial floating range are set for each time range, an optimization model taking the electricity price floating range and the electricity magnitude as decision variables is constructed by taking each time period as a basic unit, the optimal electricity price floating range and the optimal electricity magnitude selection of each time period are recorded, and an electricity price magnitude table is updated.
Further, the preliminary reference value is an optimal excitation electricity price in a segment area, and the initial floating range is a range between the excitation electricity price corresponding to an initial time node and the excitation electricity price corresponding to an end time node of the segment area; the optimization model is used for estimating the comprehensive benefit of the strategy according to the prediction result and the optimization target, reducing the electricity price floating range, and optimizing the electricity price floating range and the magnitude of each period by combining the prediction results of the characteristic identification model and the elastic estimation model.
One or more technical schemes provided by the application have at least the following technical effects or advantages:
By using the electricity price magnitude table to input a characteristic identification model and an elasticity estimation model, historical electricity price data are divided according to magnitude levels and are related to specific time nodes, and the effect of more accurate excitation electricity price control optimization is achieved by utilizing excitation electricity prices of multiple magnitudes; and by combining the electrical degree nodes with the time degree nodes, determining the span nodes to deeply analyze the load response characteristics, generating an optimized regulation report, and providing a targeted optimized regulation strategy, thereby realizing the effect of maximizing economic benefit.
Drawings
FIG. 1 is a schematic flow chart of a flexible load aggregation characteristic identification and optimization method in an embodiment of the invention;
FIG. 2 is a schematic diagram of the power price level representation in a flexible load aggregation characteristic identification and optimization method according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of a time period analysis curve in a flexible load aggregation characteristic identification and optimization method according to an embodiment of the present invention;
fig. 4 is a schematic diagram of segment region division of a flexible load aggregation characteristic identification and optimization method according to an embodiment of the present invention.
Detailed Description
In order that the application may be readily understood, a more complete description of the application will be rendered by reference to the appended drawings; the preferred embodiments of the present application are illustrated in the drawings, but the present application can be embodied in many different forms and is not limited to the embodiments described herein; rather, these embodiments are provided so that this disclosure will be thorough and complete.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs; the terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention; the term "and/or" as used herein includes any and all combinations of one or more of the associated listed items.
Embodiment one: as shown in fig. 1, a method for identifying and optimizing flexible load aggregation characteristics includes:
S100: the method comprises the steps of obtaining a characteristic identification model oriented to flexible load, wherein the input of the characteristic identification model is excitation electricity price, the output of the characteristic identification model is real-time response electricity consumption and a real-time load response characteristic vector, and the real-time load response characteristic vector comprises real-time response electricity consumption change rate, load rate fluctuation, response sensitivity and response delay.
In some embodiments, the input of the established characteristic identification model is the excitation electricity price, the output of the characteristic identification model is the real-time response electricity consumption and the real-time load response characteristic vector, the established characteristic identification-oriented neural network model is trained, specifically, the excitation electricity price and the historical response electricity consumption are utilized to form a first training data set, the loss function of the neural network model is set as a mean square error function, the first training data set is utilized to train the characteristic identification-oriented neural network model by adopting a random gradient descent algorithm, more than two hidden layers are arranged in the characteristic identification model, two output layers are arranged, for the output of the real-time response electricity consumption, corresponding to 1 neuron, the predicted real-time response electricity consumption is output; For the real-time load response feature vector output, adding a neuron for each feature, and respectively corresponding the output of each neuron to the predicted value of each feature.
In some embodiments, each time period corresponds to one excitation price, each input process selects excitation prices for a plurality of time periods, T is the total number of time periods, N is the number of features, the input layer receives T excitation prices as input, the first layer of hidden layers receives T excitation prices, outputs N 1 hidden units, uses a ReLU activation function, the second layer of hidden layers receives N 2 hidden units of the first layer of hidden layers, and one neuron is used to output predicted real-time response power applications，The neurons are respectively used for outputting predicted values of the characteristics，，......，. Setting a loss function L to comprehensively consider the real-time response power consumption and the prediction errors of all feature vectors:
Wherein, alpha and beta are weight coefficients for balancing the importance of the prediction error of the real-time response power consumption and the prediction error of the real-time load response characteristic vector; is the true value of the f feature of the t-th period, The method is characterized in that the actual real-time response electricity consumption in the t-th period is measured and acquired by setting a plurality of metering points in a power network to use an electric energy meter.
In some embodiments, an original data set is divided into a training set, a test set and a verification set, the verification set comprises diversified data samples, various situations occurring in practical application are reflected, the original data set comprises excitation electricity price and historical response electricity consumption corresponding to the excitation electricity price and historical load response feature vectors, training is carried out on a model by using training set data, model parameters are adjusted, an input layer of a neural network model receives original excitation electricity price data, the excitation electricity price data comprises a plurality of excitation electricity price data of a plurality of time periods, the electricity price change in a period is represented, the excitation electricity price data enters a first layer hidden layer and then undergoes nonlinear transformation, namely nonlinear mapping is carried out on the input data through an activating function, initial features are extracted, the initial features are trend fluctuation of the excitation electricity price in a period, nonlinear combination and transformation are carried out on output results of a previous layer in a second layer and a subsequent hidden layer, a relation structure of the excitation electricity price and the load response feature vectors is obtained, high-level features are generated, the high-level features are combined on an output layer, the neural network is used for generating the excitation electricity price and the response feature vectors, the response factors of the model response to the model is estimated by using the real-time response parameters, the real-time response parameters are estimated, the real-time response factors can be estimated, and the real-time response factors are estimated, and the model response factors can be estimated and verified.
In some embodiments, the real-time load response feature vector includes a real-time response power consumption change rate, load rate fluctuation, response sensitivity and response delay, and by acquiring and monitoring the real-time load response feature vector, the running state of the flexible load under the excitation power price can be found in time, potential supply-demand contradictions and risk hidden dangers can be accurately identified according to the real-time state and the historical data, the stability and the reliability of the flexible load are enhanced, the safety and the continuity of power supply are ensured, the optimal excitation power price and the optimal real-time response power consumption are selected under the condition of ensuring the safety supply, the cost is reduced, and the economic benefit is improved.
Specifically, the change rate of the power consumption is responded in real time）：
Wherein, AndThe actual real-time response electricity consumption of the t time period and the t-1 time period are respectively, and the real-time response electricity consumption is subjected to difference making through the real-time response electricity consumption monitored in the adjacent time period and the ratio is generated to obtain the real-time response electricity consumption change rate; load factor fluctuation [ ]）：For the maximum possible load amount for the t-th period,For the maximum possible load amount of the t-1 th period, then
The response delay (τ) needs to be obtained through experiments or long-time observation, and can be specifically adjusted according to practical situations.
Response sensitivity [ ]）：
Wherein, ，，Represents the amount of change in the amount of power consumption in response between the t-th period and the t-1 th period,Represents the electricity rate variation amount between the t-th period and the t-1 th period,Refers to the excitation electricity price at the t-th moment,Refers to the excitation electricity price at time t-1.
S200: the method comprises the steps of obtaining a flexible load-oriented electricity price magnitude list, wherein the electricity price magnitude list comprises excitation electricity price and corresponding time nodes, selecting initial excitation electricity price to be input into a characteristic identification model for the first time, sequentially selecting residual excitation electricity price to be input according to the initial excitation electricity price and magnitude rule, respectively obtaining corresponding real-time response electricity consumption and real-time load response feature vectors, and marking the time nodes corresponding to the excitation electricity price conforming to the magnitude rule as time nodes.
In some embodiments, as shown in fig. 2, the electricity price magnitude table includes all excitation electricity prices, where the excitation electricity prices are historical electricity price data, and the electricity prices are divided into the levels of element, angle, minute, li and the like, for example, from 0.00 element to several elements, and each 0.01 element or 0.1 element increase is regarded as an electricity price level; the exciting electricity price has corresponding time nodes in an electricity price magnitude table, wherein the time nodes adopt 24 hours for timing, namely, the time nodes record from 0 time to 24 hours, and the specific time of the exciting electricity price in one day is represented; when the excitation electricity price is specifically selected for input, randomly selecting one excitation electricity price as an initial excitation electricity price, traversing an electricity price magnitude table by taking the initial excitation electricity price as a reference, inquiring whether the difference between the excitation electricity price and the initial excitation electricity price meets magnitude rules or not, and inputting the excitation electricity price meeting the magnitude rules into a characteristic identification model to generate corresponding real-time response electricity consumption and real-time load response feature vectors.
In some embodiments, whether the system security constraint is satisfied is determined based on the real-time response power consumption, and the system security constraint can be read from various parameters and performance requirements of the aggregation optimization of the initial configuration; initial configuration typically includes checking the status of the communication network, importing a historical database, importing a historical empirical model, and reading various parameters and performance requirements for aggregate optimization. Based on the obtained real-time response electricity consumption, the satisfaction condition of the system safety constraint is calculated and judged by combining the expression of the system safety constraint. For example, a common system security constraint is a system capacity limitation constraint, and if the sum of all response power consumption amounts obtained based on real-time response power consumption amounts exceeds a given capacity limitation value, the system security constraint is not satisfied; otherwise, if the system safety constraint is met, if the system safety constraint has various constraints such as capacity limitation, voltage stability limitation, frequency stability limitation and the like, whether all the system safety constraints are met or not needs to be judged, and if all the system safety constraints are met, the flexible load state is indicated that the system running risk is not caused at the moment; for the excitation electricity price and the real-time response electricity consumption which do not meet the safety constraint, the iterative optimization calculation needs to be continuously operated to adjust the excitation electricity price, so that the real-time response electricity consumption of the flexible load is changed, and finally the safety constraint is met.
The magnitude rule is to select the excitation electricity price which is different in magnitude from the initial excitation electricity price or not smaller than a measurement threshold value as the excitation electricity price of the subsequent input by taking the initial excitation electricity price as a reference; and calculating the variation amplitude between the excitation electricity prices with the same magnitude, and comparing the variation amplitude with the measurement threshold value. The measurement threshold is set according to actual conditions and experience and is used for measuring the change amplitude of the excitation electricity price.
In some embodiments, the magnitude rule takes different magnitudes and measurement thresholds as dividing criteria and screening rules, for example, the initial excitation electricity price is 0.5 yuan, the initial excitation electricity price is respectively '0, 5, 0 and 0' under the sequence of 'yuan, angle, division and li' of the electricity price magnitude table, when the electricity price magnitude table is traversed subsequently, firstly, traversing the magnitude with the corresponding value of 0 from left to right, and when the excitation electricity price with the corresponding value of not 0 under the sequence of 'yuan, division and li' is traversed, the excitation electricity price with the different magnitude is obtained; the measurement thresholds are experimentally determined, the measurement thresholds with different magnitudes are differently set, the magnitude is larger, the measurement thresholds are smaller, the sensitivity of different enterprises to electricity prices is adjusted in actual use, preferably, the measurement thresholds of 'yuan, angle, minute and li' are respectively set as '1, 3, 5 and 7', and the time node corresponding to the subsequent excitation electricity price conforming to the magnitude rule is marked as a time node; when the excitation electricity price comparison is carried out, two different standards exist, and if the subsequent excitation electricity price and the initial excitation electricity price belong to different orders, the subsequent excitation electricity price is used as the excitation electricity price of the subsequent input; if the subsequent excitation electricity price and the initial excitation electricity price belong to the same magnitude, judging the variation amplitude, wherein the variation amplitude is the difference between the subsequent excitation electricity price and the initial excitation electricity price, the ratio obtained by using the absolute value of the difference and the initial excitation electricity price, and if the variation amplitude is not smaller than a measurement threshold value, the variation amplitude is used as the excitation electricity price of the subsequent input; if the measured value is smaller than the measurement threshold value, the exciting electricity price is not used as the subsequent input.
S300: the output real-time response electricity consumption rate is obtained, an electricity amplitude threshold is preset according to the historical real-time response electricity consumption rate, the real-time response electricity consumption rate is compared with the electricity amplitude threshold, and a time node corresponding to the real-time response electricity consumption which is not smaller than the electricity amplitude threshold is marked as an electricity node. The electric quantity amplitude threshold is set according to actual conditions and historical data and is used for measuring the change amplitude of the electric quantity change rate of the real-time response.
In some embodiments, the real-time response electricity consumption generated according to the initial excitation electricity price is used as a reference, the real-time response electricity consumption generated according to the initial excitation electricity price is compared with the real-time response electricity consumption generated according to the follow-up generation to obtain a change amplitude, the amplitude of the follow-up real-time response electricity consumption increased or decreased compared with the previous one is the change amplitude, the electricity amplitude threshold value is preset according to experimental determination and historical data, preferably, the electricity amplitude threshold value is set to be 10% to represent the trend of the change amplitude of the real-time response electricity consumption most, and the time node where the real-time response electricity consumption is located, which is not smaller than the electricity amplitude threshold value, is marked as an electricity node.
S400: generating a time curve according to different excitation electricity prices and time nodes, generating an electricity curve according to the real-time response electricity consumption and the electricity nodes, determining span nodes, acquiring the real-time load response feature vector of the span nodes, and generating an optimized regulation report; the span node is a node where a time node and an electricity node coincide.
In some embodiments, as shown in fig. 3, the abscissa is a node of each period, a time curve and an electric curve of the same period are overlapped in a period analysis curve, a span node is determined according to the intersection point of the time curve and the electric curve, the change condition of a load response characteristic vector on the span node is analyzed in detail, the change quantity and the change rate of each characteristic and the comparison result with the adjacent node are included, the correlation analysis result between the excitation electricity price change and the load response characteristic vector change is displayed, the correlation analysis result includes a correlation coefficient, a regression equation and the like, the typical load response mode identified on the span node is described, the mode characteristics, the related load type, the influence of existence and the like are included, and a corresponding optimized regulation report is generated, and can be presented in various forms.
S500: and carrying out aggregation optimization on the flexible load according to the optimization regulation report and updating an electricity price magnitude table.
In some embodiments, by analyzing the span node and the real-time load response feature vector, the electricity price corresponding to the time node and the change trend of the real-time power consumption can be clearly judged, and further, the flexible load is aggregated and optimized according to the change trend, so that more accurate regulation and control are realized, and the maximized economic benefit is realized under the condition of ensuring the load safety.
In the embodiment, the historical electricity price data is divided according to the quantity level by using the electricity price magnitude table to input a characteristic identification model and is related to a specific time node, and more accurate excitation electricity price control optimization is realized by using excitation electricity prices of multiple magnitudes; identifying an electricity degree node with obvious electricity quantity change by comparing the real-time response electricity quantity under different excitation electricity prices; in combination with the time nodes in the electricity price magnitude table, further determining span nodes (namely points where the electricity degree nodes coincide with the time nodes), generating a time curve and an electricity degree curve according to the nodes, and deeply analyzing real-time load response feature vectors; generating an optimized regulation report based on a real-time load response feature vector analysis result of the span node, and providing a targeted optimized regulation strategy; and dynamically adjusting an excitation electricity price strategy according to the electricity price magnitude table and the real-time system state, so as to ensure that the economic benefit is maximized on the premise of meeting the system safety constraint.
The technical scheme provided by the embodiment of the application at least has the following technical effects or advantages:
According to the application, the historical electricity price data is divided according to the quantity level by using the electricity price magnitude table to input the characteristic identification model and is related to a specific time node, and the effect of more accurate excitation electricity price control optimization is realized by utilizing the excitation electricity prices of multiple magnitudes; and determining span nodes by combining the electrical degree nodes with the time degree nodes, deeply analyzing real-time load response feature vectors, generating an optimized regulation report, and providing a targeted optimized regulation strategy, thereby realizing the effect of maximizing economic benefit.
Embodiment two: in the first embodiment, an electricity price level table is set, a time node and an electricity node are obtained, a span node is obtained and analyzed to obtain an optimized regulation report for accurate regulation, and the embodiment is further improved on the basis of the embodiment.
Acquiring two adjacent span nodes, dividing a time region between the two span nodes into a plurality of auxiliary control regions according to excitation electricity prices, wherein the same excitation electricity price is the same auxiliary control region; in each auxiliary control area, dividing the auxiliary control area into a plurality of segment areas according to the difference value of the exciting electricity price; the segment areas correspond to a newly divided excitation electricity price, and corresponding real-time electricity consumption and real-time load response characteristic vectors are obtained according to the new excitation electricity price; and generating an optimized regulation report according to the optimal excitation electricity price corresponding to the fragment area and the corresponding time range obtained by analyzing the real-time response electricity consumption and the real-time load response characteristic vector.
The difference value is the difference value of excitation electricity prices among different auxiliary control areas, and is obtained by recording the calculation of the ending electricity price and the starting electricity price of the auxiliary control areas, and the auxiliary control areas with the difference value being more than 0, namely the ending electricity price being more than the starting electricity price, are divided; and setting an auxiliary control area with the difference value smaller than 0, namely the ending electricity price smaller than the starting electricity price as a pending area.
In some embodiments, according to two adjacent span nodes, dividing a time area between the two span nodes into a plurality of auxiliary control areas according to excitation electricity prices, wherein the same excitation electricity price is the same auxiliary control area, and the number of the auxiliary control areas is consistent with the number of different excitation electricity prices; dividing the auxiliary control area into a plurality of segment areas according to the difference value of the excitation electricity price, wherein the difference value is the difference value of the excitation electricity price between different auxiliary control areas, performing difference calculation by using the ending electricity price and the starting electricity price of the auxiliary control area, and dividing the auxiliary control area with the difference value being greater than 0, namely the ending electricity price being greater than the starting electricity price; and (3) for the auxiliary control area with the difference value smaller than 0, namely the ending electricity price smaller than the starting electricity price, not performing further division, preferably, setting the auxiliary control area which is not subjected to division as a pending area, wherein the pending area is used for re-dividing the auxiliary control area after the processing of the rest auxiliary control areas is finished.
The segment areas are divided according to the next magnitude of the difference, the maximum number of the segment areas is not more than 10, new excitation electricity prices are set for each segment area after the division is completed, the time length of the auxiliary control area is divided according to the number of the segment areas, and the initial time node and the end time node of each segment area are recorded.
In some embodiments, the auxiliary control areas are divided according to the next magnitude of the difference, the maximum number of the auxiliary control areas is not more than 10, the excitation electricity price corresponding to the first segment area is the initial electricity price of the auxiliary control area, the excitation electricity price corresponding to the last segment area is the end electricity price of the auxiliary control area, for each segment area, the excitation electricity price is gradually increased according to the set increasing magnitude from the initial electricity price, for example, the initial electricity price of the auxiliary control area is 0.5, the end electricity price is 0.6, the magnitude is an angle, the auxiliary control area is divided according to the new magnitude, namely, the auxiliary control area is divided according to the new magnitude, the auxiliary control area is divided into ten segment areas, namely, 0.5, 0.51, 0.52 and the auxiliary control area, the time of each segment area is set according to the magnitude of the auxiliary control area after division, for example, the time of the auxiliary control area is small, the auxiliary control area is divided into the corresponding segment areas according to the initial electricity price of 10, the response time of each segment is set to the corresponding segment area, the response time of the corresponding segment area is set to the initial power value of the real-time vector, and the real-time vector is generated according to the real-time response time vector of the response of the corresponding segment area, and the real-time vector is analyzed according to the real-time vector of the set to the initial electricity price of the power consumption.
In some embodiments, according to the real-time response electricity consumption and the real-time load response feature vector obtained from each segment region, according to the change of the excitation electricity price and the real-time response electricity consumption, the load response conditions under different excitation electricity prices are recorded, the difference of the load response between adjacent segment regions is compared, the influence of the increase amplitude on the load response is evaluated, and is compared with the real-time response electricity consumption between span nodes, the load response rules under different excitation electricity price increase amplitudes are analyzed, an aggregation optimization regulation strategy for flexible loads is provided based on the test result, the style response characteristics of different auxiliary control regions and segment regions are analyzed, and differential regulation suggestions are provided, the regulation suggestions comprise regulation concrete strategies, expected effects, risk evaluations and the like, according to the optimization regulation suggestions, the excitation electricity price setting in an electricity price table is adjusted, and the effective excitation electricity price increase amplitude found in the test is included in the electricity price magnitude table so as to optimize the load response effect.
In some embodiments, the optimal excitation electricity price of the segment area and the time range are obtained, an optimal regulation strategy is set, different optimal excitation electricity prices are used in different time ranges, further, optimal real-time response electricity consumption is obtained, accurate regulation and control of the excitation electricity price are achieved on the premise of ensuring load safety, and the maximum economic benefit is obtained.
In the embodiment, a method for dividing auxiliary control areas and segment areas is introduced, and the time areas between two adjacent span nodes are divided into a plurality of auxiliary control areas according to the excitation electricity price, and the auxiliary control areas are further divided into the segment areas according to the electricity price difference value, so that finer regulation and control of the time areas are realized; in each segment area, dynamic electricity price test is implemented by setting new excitation electricity price and obtaining corresponding real-time response electricity consumption and real-time load response characteristic vector, load response is observed under different electricity price levels, and the optimal excitation electricity price in the segment area is found.
The technical scheme provided by the embodiment of the application at least has the following technical effects or advantages:
According to the application, through dividing the auxiliary control area and the segment area and analyzing the load response under different excitation electricity prices in each segment area, the effect of more finely regulating and controlling the flexible load is realized; the optimal excitation electricity price and the corresponding time range are obtained based on data analysis, so that the electricity utilization behavior of the flexible load can be guided more effectively, the configuration and the use efficiency of the power resource are optimized, and the overall economic benefit is improved.
Embodiment III: in the above embodiment, the time length between two adjacent span nodes is divided into a plurality of auxiliary control areas and segment areas, but the time range of each segment area is uniform, which affects the accurate regulation and control.
And acquiring an initial time node of the segment area, gradually increasing a time range according to an extension step length, acquiring the real-time response electricity consumption change rate and the real-time load response characteristic vector according to the current time range and the new excitation electricity price, judging whether a preset condition is reached, stopping extension if the preset condition is reached or the maximum extension times are reached, and recording the time node at the moment as an end time node.
The maximum extension times are preset according to the time length and the extension step length of the auxiliary control area.
The preset condition comprises a first threshold value and a second threshold value, and the first threshold value is set according to the real-time response power consumption change rate; the second threshold is set according to a comprehensive fluctuation value, the comprehensive fluctuation value is generated comprehensively according to the load rate fluctuation, the response sensitivity and the response delay, and if the real-time response electricity consumption change rate reaches the first threshold or the comprehensive fluctuation value reaches the second threshold, the extension is stopped.
In some embodiments, the extension step is preset, the extension step being the length of time that increases with each extension, such as three minutes, five minutes, etc.; setting the extension step length according to the time length of the auxiliary control area, setting different extension step lengths according to the current electricity price magnitude information of the auxiliary control area and different magnitudes, wherein the larger the magnitude is, the smaller the extension step length is, adjusting according to actual conditions, setting the maximum extension times according to the extension step length and the time length of the auxiliary control area, and the time length of the auxiliary control area is not exceeded; and obtaining a new time range by adding a step length every time, predicting the real-time response power consumption and the real-time load response characteristic vector by using a characteristic identification model according to the new excitation power price, judging whether a preset condition is met, if not, continuing to extend, expanding the time range, if the preset condition is met or the maximum extension times are reached, stopping extending, determining the current time node as an end time node, and generating the time range from the initial time node to the end time node.
In some embodiments, as shown in fig. 4, the abscissa is a time period node, a first threshold and a second threshold are set between span nodes, that is, between T1-T2, the actual curve represents a specific change in the extension process, the extension is stopped when the first threshold is reached at T1, the time period T1-T1 is the time range of a segment region, the corresponding excitation electricity price at T1 is the optimal excitation electricity price of the time range, and T1-T2 and T2-T2 are the time ranges of the generated second and third segment regions, and fig. 4 is only a schematic diagram; the preset condition is provided with a first threshold value and a second threshold value, the first threshold value is set according to the real-time response electricity consumption change rate, the real-time response electricity consumption change rate can rapidly capture the real-time change of the load, the transverse and longitudinal comparison between different time periods and different load types is convenient, and when the real-time response electricity consumption change rate reaches the first threshold value, the time range has obvious change trend; the second threshold is set according to a real-time load response feature vector, a plurality of key indexes exist in the real-time load response feature vector, the real-time response electricity consumption change rate is used as a first threshold, a comprehensive fluctuation value is set as a second threshold, the comprehensive fluctuation value is comprehensively generated according to the load rate fluctuation, the response sensitivity and the response delay, the safety of the electricity consumption load is measured, the stability of the load is ensured when the step length is extended, when the second threshold is set, a reasonable change range of each key index is needed to be obtained according to historical data information analysis, a reasonable load rate fluctuation range is needed to be set for load rate fluctuation, for example, a maximum allowable fluctuation percentage is set, for example, 5%, and when the real-time monitoring finds that the real-time load rate fluctuation exceeds the maximum allowable fluctuation percentage, the load stability is judged to be threatened at the moment; aiming at two key indexes of which the response sensitivity and the response delay are relative, analyzing the sensitivity degree of different load types to the change of electricity price, setting a sensitivity upper limit, and avoiding the severe change of load caused by excessive sensitivity; the response delay can not directly reflect the stability, but too high or too short response delay can not accurately identify the change condition of the current excitation electricity price, and the minimum delay time is set to ensure that the change condition of the excitation electricity price can be accurately identified; the second threshold is set by combining three indexes of load rate fluctuation, response sensitivity and response delay, specifically three index data of load rate fluctuation range, sensitivity upper limit and minimum delay time, and is obtained by carrying out normalization processing on the three index data and dividing different weight values for weighted summation; and according to the initial time node of the segment area, automatically dividing the time range according to the set extension step length and preset conditions.
The technical scheme provided by the embodiment of the application at least has the following technical effects or advantages:
The application avoids inaccuracy possibly brought by time range uniform division in a dynamic extension mode, so that the time range is divided more flexibly, and the application can be better adapted to the change of load response characteristics; by monitoring the real-time response electricity consumption change rate and the real-time load response feature vector and stopping extending when the preset condition is reached, each segment area is ensured to accurately reflect the response characteristic of the load in the time period, and accurate optimization regulation and control are realized.
Embodiment four: in the above embodiment, the time range of the segment area is automatically obtained according to the preset conditions and the set extension step length, so as to realize more accurate optimization and regulation, and the present embodiment is further limited on the basis of the above embodiment.
And acquiring the time range of the segment area and the excitation electricity price, setting a preliminary reference value and an initial floating range for each time range, constructing an optimization model taking the floating range and the magnitude of the electricity price as decision variables by taking each time period as a basic unit, recording the optimal floating range and the optimal magnitude selection of the electricity price in each time period, and updating an electricity price magnitude table.
The initial reference value is the optimal excitation electricity price in the segment area, and the initial floating range is the range between the excitation electricity price corresponding to the initial time node and the excitation electricity price corresponding to the end time node of the segment area; the optimization model is used for evaluating the comprehensive benefit of the strategy according to the prediction result and the optimization target, reducing the electricity price floating range and optimizing the electricity price floating range and the magnitude of each period according to the prediction result of the characteristic identification model.
In some embodiments, the maintenance length and the optimal excitation electricity price of the segment area can be accurately obtained according to the time range of the segment area and the excitation electricity price, each time range comprises a plurality of time periods, a preliminary reference value and an initial floating range are set in the time range, the initial floating range refers to the floating range of the excitation electricity price, as shown in fig. 4, the time period T1-T1 is the time range of one segment area, the excitation electricity price corresponding to T1 is the optimal excitation electricity price of the time range, and the excitation electricity price corresponding to T1 are the initial floating range; constructing an optimization model taking a power price floating range and an order as decision variables by taking each time interval as a basis unit, optimizing the power price floating range and the order of each time interval according to a prediction result of a characteristic identification model by adopting heuristic algorithms such as a genetic algorithm and a particle swarm optimization algorithm, generating a series of candidate power price strategies according to initial power price reference values and floating ranges in each time interval, predicting load responses under different candidate power price strategies by using the characteristic identification model, evaluating comprehensive benefits, especially economic benefits, of each candidate strategy according to the prediction result and an optimization target, gradually reducing the power price floating range until an optimal solution is found, recording the optimal power price floating range and the optimal order selection of each time interval, and updating a power price table; integrating the optimal electricity price floating range and the optimal magnitude selection of each period, compiling a new electricity price magnitude table according to the integrated optimization result, determining the electricity price floating range and the optimal magnitude of each period, and carrying out safety constraint verification on the new electricity price magnitude table to ensure that the new electricity price magnitude table meets the safety standard of the system; and implementing a new electricity price strategy, and adjusting the excitation electricity price of each period according to the new electricity price magnitude table.
In this embodiment, based on the dynamic time range dividing method in the previous embodiment, the time range of the segment area and the corresponding excitation electricity price are obtained, the time range of each segment area is set with a preliminary reference value (i.e., the optimal excitation electricity price of the area) and an initial floating range (determined by the excitation electricity price of the area starting and ending time nodes), an optimization model with the electricity price floating range and the magnitude as decision variables is constructed based on each time period, the optimal electricity price floating range and the optimal magnitude of each time period are integrated, the electricity price magnitude table is updated, and the electricity price floating range and the optimal magnitude of each time period are defined.
The technical scheme provided by the embodiment of the application at least has the following technical effects or advantages:
According to the application, the optimal electricity price floating range and the optimal magnitude are set for each time period, so that the electricity price regulation and control accuracy is realized, the electricity price strategy can be adjusted in real time according to the load characteristics and the system requirements, the flexibility of the strategy is enhanced, the optimal electricity price strategy of each time period is found, and the effect of improving the economic benefit is realized.
The above description is only of the preferred embodiments of the present invention and is not intended to limit the present invention, but various modifications and variations can be made to the present invention by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411139397.2A CN118676940B (en) | 2024-08-20 | 2024-08-20 | Flexible load aggregation characteristic identification and optimization method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411139397.2A CN118676940B (en) | 2024-08-20 | 2024-08-20 | Flexible load aggregation characteristic identification and optimization method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118676940A true CN118676940A (en) | 2024-09-20 |
| CN118676940B CN118676940B (en) | 2024-11-08 |
Family
ID=92724830
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202411139397.2A Active CN118676940B (en) | 2024-08-20 | 2024-08-20 | Flexible load aggregation characteristic identification and optimization method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118676940B (en) |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180357730A1 (en) * | 2017-06-12 | 2018-12-13 | Tata Consultancy Services Limited | Systems and methods for optimizing incentives for demand response |
| CN113610351A (en) * | 2021-07-13 | 2021-11-05 | 国网浙江省电力有限公司电力科学研究院 | User demand response capability assessment method, system, terminal and medium |
| WO2021244000A1 (en) * | 2020-06-03 | 2021-12-09 | 国网上海市电力公司 | Virtual aggregation system and method for regional energy source complex |
| CN114154900A (en) * | 2021-12-09 | 2022-03-08 | 国网电子商务有限公司 | Power supply control method and device |
| WO2022135066A1 (en) * | 2020-12-25 | 2022-06-30 | 南京理工大学 | Temporal difference-based hybrid flow-shop scheduling method |
| CN115358640A (en) * | 2022-10-20 | 2022-11-18 | 四川中电启明星信息技术有限公司 | Residential district demand response method based on stepped excitation |
| CN115496627A (en) * | 2022-11-22 | 2022-12-20 | 国网天津市电力公司电力科学研究院 | A Method and System for Evaluating the Response Potential of Adjustable Resources |
| CN115629576A (en) * | 2022-09-16 | 2023-01-20 | 清华大学 | Non-invasive flexible load aggregation characteristic identification and optimization method, device and equipment |
| CN115994656A (en) * | 2022-12-07 | 2023-04-21 | 国网河南省电力公司经济技术研究院 | Virtual power plant economic dispatching method considering excitation demand response under time-of-use electricity price |
| CN116960944A (en) * | 2023-06-28 | 2023-10-27 | 淮阴工学院 | Excitation type flexible compensation electricity price method based on credit |
| WO2024056050A1 (en) * | 2022-09-15 | 2024-03-21 | 清华大学 | Massive flexible load rapid aggregation control method and apparatus |
-
2024
- 2024-08-20 CN CN202411139397.2A patent/CN118676940B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180357730A1 (en) * | 2017-06-12 | 2018-12-13 | Tata Consultancy Services Limited | Systems and methods for optimizing incentives for demand response |
| WO2021244000A1 (en) * | 2020-06-03 | 2021-12-09 | 国网上海市电力公司 | Virtual aggregation system and method for regional energy source complex |
| WO2022135066A1 (en) * | 2020-12-25 | 2022-06-30 | 南京理工大学 | Temporal difference-based hybrid flow-shop scheduling method |
| CN113610351A (en) * | 2021-07-13 | 2021-11-05 | 国网浙江省电力有限公司电力科学研究院 | User demand response capability assessment method, system, terminal and medium |
| CN114154900A (en) * | 2021-12-09 | 2022-03-08 | 国网电子商务有限公司 | Power supply control method and device |
| WO2024056050A1 (en) * | 2022-09-15 | 2024-03-21 | 清华大学 | Massive flexible load rapid aggregation control method and apparatus |
| CN115629576A (en) * | 2022-09-16 | 2023-01-20 | 清华大学 | Non-invasive flexible load aggregation characteristic identification and optimization method, device and equipment |
| CN115358640A (en) * | 2022-10-20 | 2022-11-18 | 四川中电启明星信息技术有限公司 | Residential district demand response method based on stepped excitation |
| CN115496627A (en) * | 2022-11-22 | 2022-12-20 | 国网天津市电力公司电力科学研究院 | A Method and System for Evaluating the Response Potential of Adjustable Resources |
| CN115994656A (en) * | 2022-12-07 | 2023-04-21 | 国网河南省电力公司经济技术研究院 | Virtual power plant economic dispatching method considering excitation demand response under time-of-use electricity price |
| CN116960944A (en) * | 2023-06-28 | 2023-10-27 | 淮阴工学院 | Excitation type flexible compensation electricity price method based on credit |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118676940B (en) | 2024-11-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109659933B (en) | Electric energy quality prediction method for power distribution network with distributed power supply based on deep learning model | |
| Liu et al. | A hybrid FCW-EMD and KF-BA-SVM based model for short-term load forecasting | |
| Saber et al. | Short term load forecasting using multiple linear regression for big data | |
| Osman et al. | Neural network based approach for short-term load forecasting | |
| Wang et al. | Hybrid interval AHP-entropy method for electricity user evaluation in smart electricity utilization | |
| Weng et al. | Probabilistic baseline estimation via gaussian process | |
| Kebriaei et al. | Short-term load forecasting with a new nonsymmetric penalty function | |
| CN110163429B (en) | A Short-Term Load Forecasting Method Based on Optimal Screening of Similar Days | |
| Ghasempour et al. | Short-term electric load prediction in smart grid using multi-output Gaussian processes regression | |
| Khalina et al. | Formation of organizational support for the management of the economic security of engineering enterprises: methodical and practical aspects | |
| CN113408869A (en) | Power distribution network construction target risk assessment method | |
| Hoiles et al. | Nonparametric demand forecasting and detection of energy aware consumers | |
| CN104123617A (en) | A Method of Power Load Forecasting | |
| Zhang et al. | Soft actor–critic algorithm featured residential demand response strategic bidding for load aggregators | |
| Li et al. | Day-ahead risk averse market clearing considering demand response with data-driven load uncertainty representation: A Singapore electricity market study | |
| CN105224997A (en) | A kind of planning load forecasting method containing multiple uncertain factor impact | |
| CN109447332A (en) | A kind of Middle-long Electric Power Load Forecast method suitable for S type load curve | |
| CN115358476A (en) | Demand response potential prediction method, demand response potential prediction device, computer equipment and storage medium | |
| CN119160030A (en) | Intelligent energy-saving management method and system for charging piles based on big data | |
| Zheng et al. | An application of machine learning for a smart grid resource allocation problem | |
| CN118676940B (en) | Flexible load aggregation characteristic identification and optimization method | |
| Sari et al. | The effectiveness of hybrid backpropagation Neural Network model and TSK Fuzzy Inference System for inflation forecasting | |
| CN117745119A (en) | Power transmission and transformation project construction bearing capacity early warning method and device | |
| Gajowniczek et al. | Seasonal peak demand classification with machine learning techniques | |
| CN117691594B (en) | Energy saving and consumption reduction judging method and system for transformer |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |