User adjustable load evaluation method and system oriented to demand response

Abstract

The invention relates to the technical field of power information processing, and provides a demand response-oriented user adjustable load evaluation method and system. The method comprises the following steps: based on acquired user load data before a demand response day, meteorological data of the demand response day and holiday and festival data, acquiring a trained user short-term power consumption load prediction model to obtain a user power consumption load prediction amount of the demand response day; calculating the real load response amount of the demand response day according to the difference value between the user electricity real load amount of the demand response day and the user electricity load predicted amount; calculating the performance responsivity of the user according to the ratio of the real load response amount of the demand response day to the reported expected response amount; obtaining the performance responsivity of the same user in the application scenarios of different time periods, different electricity price differences and different dates; and based on the application scene, matching the minimum value of the performance responsivity of the user demand response, and in combination with the user demand response filling amount, evaluating the adjustable resource potential of the user side.

Description

Technical Field

[0001] The present invention relates to the technical field of power information processing, and in particular, to a method and system for evaluating user adjustable load for demand response. Background Art

[0002] The statements in this part only provide background technical information related to the present invention and do not necessarily constitute prior art.

[0003] Demand response is an act of guiding power users to voluntarily change their electricity consumption patterns according to power market price signals or incentive measures, reducing electricity consumption during peak periods of the system, thereby promoting the stable operation of the power system. With the rapid economic development and continuous population growth, the power demand continues to climb. Especially during peak periods, the power supply is tight, and the power supply and demand are seriously unbalanced. Demand response has become an important measure to solve the power supply and demand imbalance. With the advancement of the power market reform, more and more power generation enterprises, distributed energy enterprises, energy-saving service companies, etc. have gradually participated in the demand response market, forming diversified implementation entities. However, due to the increase in participating users, the performance capabilities are uneven.

[0004] Currently, the sample data of demand response user cases is relatively small, making it difficult to directly model and accurately evaluate the user's performance response volume. To address this problem, the existing technology evaluates the performance capabilities of users' power demand response by establishing a user load prediction model. However, this method has problems such as difficulty in capturing long-term time series dependence relationships and difficulty in taking into account the short-term time series dynamic fluctuations within a day, resulting in problems such as high model training difficulty and low computational efficiency. This leads to the inability to accurately evaluate the performance capabilities of users' power demand response, insufficient understanding of users' electricity consumption behaviors and demand response potentials, which is not conducive to power system operators reasonably allocating power resources, and the power resource utilization rate needs to be further improved. Summary of the Invention

[0005] To solve the technical problems existing in the above background art, the present invention provides a method and system for evaluating user adjustable load for demand response. The present invention fully considers the performance influencing factors in the user demand response process, constructs a user short-term electricity load prediction model using a dual-branch structure, realizes the prediction of the normal load volume on the demand response day of users, and obtains the performance response degrees of users in different scenarios by calculating the ratio of the user's actual load response volume to the reported response volume, thereby realizing the evaluation of the potential of adjustable resources on the user side, expanding the market for high-potential users not included in the adjustable resource library, thereby expanding the adjustable resource library, enhancing the grid regulation ability, and improving the power resource utilization rate.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] The first aspect of the present invention provides a method for evaluating user-adjustable loads for demand response.

[0008] A method for evaluating user-adjustable loads for demand response includes:

[0009] Based on the obtained user load data before the demand response day, the meteorological data on the demand response day, and the holiday data, using the trained user short-term electricity load prediction model to obtain the predicted user electricity load on the demand response day;

[0010] According to the difference between the actual user electricity load on the demand response day and the predicted user electricity load, calculate the actual load response on the demand response day; according to the ratio between the actual load response on the demand response day and the reported expected response amount, calculate the compliance response degree of the user; based on the user's historical demand response data, calculate the compliance response degree for each time, and obtain the compliance response degree of the same user under application scenarios with different time periods, different electricity price differences, and different dates;

[0011] Based on the minimum value of the compliance response degree of the user demand response matched by the application scenario, combined with the reported amount of the user demand response, evaluate the potential of adjustable resources on the user side.

[0012] Further, the process of using the user short-term electricity load prediction model includes:

[0013] Convert the obtained user load data before the demand response day, the meteorological data on the demand response day, and the holiday data into two-dimensional data, respectively model the short-term intra-day time series and the long-term daily periodic pattern, and perform normalization processing;

[0014] Based on the two-dimensional data, use a dual-branch structure to capture information. Capture long-term time series features through the long-term time series feature extraction branch, and capture short-term intra-day time series features through the short-term intra-day time series feature extraction branch;

[0015] Fuse the long-term time series features and the short-term intra-day time series features, and through linear processing, obtain the predicted user electricity load on the demand response day.

[0016] Further, the long-term time series feature extraction branch adopts a network composed of TCN, GRU, and an attention mechanism, and the short-term intra-day time series feature extraction branch adopts a linear layer.

[0017] Further, the method for capturing long-term time series features through the long-term time series feature extraction branch includes: using TCN to extract the long-term historical information features of each time step, introducing an attention mechanism to automatically screen the important information features in the long-term historical information features of each time step, and calculating the weights of the important information features; inputting the important information features into GRU for parallel information selection and fusion to obtain long-term time series features.

[0018] Further, the intraday short-term time series feature extraction branch captures intraday short-term time series features; the method includes: aggregating intraday short-term fluctuation data using a linear layer to obtain intraday short-term time series features.

[0019] Further, based on the minimum value of the user demand response compliance response degree matched by the application scenario, combined with the user demand response reporting quantity, the potential of the user-side adjustable resources is evaluated; the method includes: based on the minimum value of the user demand response compliance response degree matched by various application scenarios, combined with the user demand response reporting quantity, calculating the evaluation value of the user-side adjustable resource potential, and based on the evaluation of the compliance ability of the user demand response, the demand side formulates a high-potential and compliance white list, and gives priority to the users on the list when conducting demand response invitations.

[0020] The second aspect of the present invention provides a user adjustable load evaluation system for demand response.

[0021] A user adjustable load evaluation system for demand response includes:

[0022] A prediction module, which is configured to: based on the acquired user load data before the demand response day, the meteorological data and holiday data on the demand response day, and using the trained user short-term electricity load prediction model, obtain the predicted user electricity load on the demand response day;

[0023] A compliance response degree calculation module, which is configured to: calculate the actual load response amount on the demand response day according to the difference between the actual user electricity load amount and the predicted user electricity load amount on the demand response day; calculate the compliance response degree of the user according to the ratio between the actual load response amount and the reported expected response amount on the demand response day; calculate the compliance response degrees of previous times based on the user's historical demand response data, and obtain the compliance response degrees of the same user under different time periods, different electricity price differences, and different application scenarios on different dates;

[0024] An evaluation module, which is configured to: based on the minimum value of the user demand response compliance response degree matched by the application scenario, combined with the user demand response reporting quantity, evaluate the potential of the user-side adjustable resources.

[0025] The third aspect of the present invention provides a computer-readable storage medium.

[0026] A computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, it realizes the steps in the user adjustable load evaluation method for demand response described in the first aspect above.

[0027] The fourth aspect of the present invention provides a computer device.

[0028] A computer device includes a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, it implements the steps in the demand response-oriented user-adjustable load assessment method described in the first aspect above.

[0029] The fifth aspect of the present invention provides a computer program product or a computer program.

[0030] The present invention provides a computer program product or a computer program. The computer program product or the computer program includes computer instructions stored in a computer-readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, causing the computer device to execute the steps in the demand response-oriented user-adjustable load assessment method described in the first aspect above.

[0031] Compared with the prior art, the beneficial effects of the present invention are:

[0032] The present invention provides a demand response-oriented user-adjustable load assessment method and system. By obtaining the historical load data of industrial and commercial users and the data during the demand response process, classifying and preprocessing them, an initial power consumption load sample set is constructed. For different industrial and commercial user classes, based on the historical load data of users, meteorological data, and holiday data, a dual-branch structure is used to predict the next-day load data of users. Through feedback parameter adjustment, a short-term user power consumption load prediction model is obtained. Based on the short-term user power consumption load prediction model, by inputting the user load data before the demand response day, the meteorological data and holiday data on the demand response day, the load data of users on the demand response day is predicted to obtain the predicted user power consumption load. Based on the load reduction amount during the demand response process and the expected response load amount filled in by the user, the ratio of the actual load response amount of the user to the filled response amount is calculated as the compliance response degree. Based on the historical demand response data of the user, the compliance response degrees for each time are calculated to obtain the compliance response degrees of the same user under different time periods, different electricity price differences, and different date application scenarios. Based on the minimum value of the compliance response degree of the user demand response matched by the application scenario, combined with the filled amount of the user demand response, the potential of the adjustable resources on the user side is evaluated, thereby improving the accuracy of grid regulation on the user side and ensuring the balance between power grid supply and demand.

[0033] The present invention uses TCN + GRU + attention mechanism to construct a short-term user power consumption load prediction model, that is, uses a temporal convolutional network to perform parallel computing and processing on long-term historical time series data, and screens important influencing factors through the attention mechanism. The gating mechanism controls the memory matching selection and fusion in the time series data. A linear layer is used to extract the short-term intra-day fluctuation data, and the long-term and short-term data are aggregated to improve the accuracy of the user load prediction amount. At the same time, parallel data processing can effectively improve the data processing speed.

[0034] The present invention evaluates the user's demand response fulfillment ability based on specific scenarios, that is, calculates the actual load response amount on the demand response day based on the daily predicted load amount obtained from the short-term user load prediction model and the actual load amount on the demand response day, and then calculates the ratio with the reported load response amount before the demand response day, and calculates the fulfillment response degrees of different users in different scenarios, so as to accurately evaluate the user's demand response fulfillment amount. Brief Description of the Drawings

[0035] The accompanying drawings forming a part of the present invention are used to provide a further understanding of the present invention. The illustrative embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation to the present invention.

[0036] Figure 1 is a flowchart of the user adjustable load evaluation method for demand response shown by the present invention;

[0037] Figure 2 is a flowchart of the construction of the user short-term electricity load prediction model shown by the present invention;

[0038] Figure 3 is a flowchart of the calculation of the fulfillment response degree shown by the present invention;

[0039] Figure 4 is a structural diagram of the user short-term electricity load prediction model shown by the present invention;

[0040] Figure 5 is a structural diagram of the TCN+GRU+attention mechanism neural network shown by the present invention;

[0041] Figure 6 is a comparison chart of the experimental execution times of four methods shown by the present invention. Detailed Embodiments

[0042] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0043] It should be noted that the following detailed description is illustrative and is intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present invention belongs.

[0044] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0045] It should be noted that the flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of methods and systems according to various embodiments of the present disclosure. It should be noted that each block in the flowchart or block diagram may represent a module, a program segment, or a part of code, and the module, program segment, or part of code may include one or more executable instructions for implementing the logical functions specified in each embodiment. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than that marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, or they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the flowchart and / or block diagram, as well as the combination of blocks in the flowchart and / or block diagram, may be implemented using a dedicated hardware-based system for performing the specified functions or operations, or may be implemented using a combination of dedicated hardware and computer instructions.

[0046] Embodiment 1

[0047] As Figure 1 shown, this embodiment provides a method for evaluating user-adjustable loads for demand response. This embodiment takes the application of this method to a server as an example. It can be understood that this method can also be applied to a terminal, and can also be applied to a system including a terminal and a server, and is implemented through the interaction between the terminal and the server. The server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, web servers, cloud communications, middleware services, domain name services, security services CDN, and big data and artificial intelligence platforms. The terminal can be a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart watch, etc., but is not limited thereto. The terminal and the server can be directly or indirectly connected through wired or wireless communication methods, and this application does not make any restrictions here. In this embodiment, the method includes the following steps:

[0048] Step S1: Obtain the historical load data of industrial and commercial users and the data participating in the demand response process, classify and preprocess them, and construct an initial power load sample set;

[0049] Among them, the categories of industrial and commercial users are shown in Table 1:

[0050] Table 1 Classification of Industrial and Commercial Users

[0051]

[0052] Step S2: For different industrial and commercial user classes, based on the user's historical load data, meteorological data, and holiday data, use a dual-branch structure to predict the user's next-day load data. Through feedback parameter tuning, obtain the user's short-term electricity load prediction model;

[0053] As Figure 2 , Figure 3 shown, in step S2, for different electricity user classes, based on the user's historical load data, meteorological data, and holiday data, use a dual-branch structure to predict the user's next-day load data. Through feedback parameter tuning, obtain the user's short-term electricity load prediction model. The specific process is as follows:

[0054] Step S21: Input preparation stage: Reshape the input user's historical load data, meteorological data, and holiday data information from one-dimensional to two-dimensional, respectively model the information in the rows (short-term changes) and columns (long-term periodic patterns), and perform normalization processing. The specific process is as follows:

[0055]

[0056] Among them, μ X represents the mean of the input load data, represents the variance of the input load data, L h represents the number of input load data; norm is a hyperparameter that divides the data set from a one-dimensional data sequence into a two-dimensional data sequence. represents the normalized load sequence, C represents the number of load sequences; Reshape(·) represents the reshaping operation, [·] represents the concatenation operation, represents the external time features, C time represents the number of external time features.

[0057] Step S22: Input the reshaped user's historical load data, meteorological data, and holiday data information Use the dual-branch structure to capture information, capture the long-term time series features of the samples through the TCN + GRU + attention mechanism, and capture the short-term time series features within the day through the linear layer.

[0058] As Figure 5 shown, use TCN to extract the long-term historical information features of each time step, introduce the attention mechanism to automatically learn and extract the important features in the features, and then select and fuse the current and historical information of each time step through the gating mechanism to maximize the retention of the long-term periodic data features.

[0059] For the R dimension of the input data , the process of feature extraction using the convolution kernel f is as follows:

[0060]

[0061] In the formula: Y(t) represents the extracted information feature; f is the convolution kernel, ω is the weight of the convolution kernel; d is the dilation rate; x t-(d·f) represents the user's historical data of the t-th element and before t.

[0062] At this time, by introducing the attention mechanism to screen important information features, that is, calculating the importance weight α for each data feature:

[0063]

[0064] where Q, K, and V represent Query, Key, and Value respectively, and l k represents the vector length.

[0065] Input the data features with weight scoring based on the attention mechanism into the GRU gating mechanism for parallel information selection and fusion:

[0066] G = σ(W g [X, αY] + b g )

[0067] O = tanh(W t [X, αY] + b t )

[0068] where are the weight matrix and bias vector, G and O are intermediate variables, and σ(.) and tanh(.) are activation functions.

[0069] Perform an element-wise product operation on the fused information to obtain the result of the long-term time series feature extraction branch:

[0070] Out long = GeY + (1 - G)eO

[0071] where e is the element-wise product.

[0072] When extracting the intraday short-term fluctuation feature branch, use a linear layer to aggregate the intraday short-term fluctuation data:

[0073]

[0074] where represents the linear aggregation of short-term information along the P dimension to obtain the output of the intraday short-term time series branch with the same dimension as the output of the long-term time series feature extraction branch represents that through this linear layer, is fused into the global information of the sequence

[0075] The present invention constructs a short-term user electricity load prediction model by using TCN + GRU + attention mechanism, that is, using a temporal convolutional network to perform parallel computing and processing on long-term historical time series data, screening important influencing factors through the attention mechanism, and using a gating mechanism to control the memory matching selection and fusion in the time series data. A linear layer is used to extract the short-term intra-day fluctuation data, and the long-term and short-term data are aggregated to improve the accuracy of the user load prediction amount. At the same time, parallel data processing can effectively improve the data processing speed.

[0076] Step S23: Prediction stage: Merge the feature representation data of the long-term time series feature extraction branch and the short-term intra-day time series feature extraction branch, and obtain the prediction value through a linear layer.

[0077]

[0078] Among them, the output dimension obtained through the Linear(·) operation is L f = R f ×P, L f represents the length of the prediction sequence. Finally, the above output is reshaped into one dimension through the Reshape(·) operation to obtain the output result

[0079]

[0080] Step S24: Continuously feedback and adjust the parameters based on the prediction model, and finally output the short-term user electricity load prediction model.

[0081] Step S3: Based on the short-term user electricity load prediction model, input the user load data before the demand response day, the meteorological data and holiday data on the demand response day, predict the user load data on the demand response day, and obtain the user electricity load prediction amount;

[0082] Input the normal load data of the user in the n days before the user demand response into the short-term user electricity load prediction model output in step B, predict the normal load amount on the user demand response day, and output the prediction data Out.

[0083] Step S4: Based on the load reduction amount during the demand response process and the expected response load amount filled in by the user, calculate the ratio of the user's actual load response amount to the filled-in response amount as the compliance response degree;

[0084] As Figure 3 shown, step 4 calculates the ratio of the user's actual load response amount to the filled-in response amount as the compliance response degree based on the load reduction amount during the demand response process and the expected response load amount filled in by the user. The specific process is as follows:

[0085] Step S41: Set the three time nodes of before the demand response day, on the response day, and after the response day as the (T - 1)th day, the Tth day, and the (T + 1)th day respectively, and obtain the load data of the user on the (T - 1)th day, the Tth day, and the (T + 1)th day respectively;

[0086] On the (T - 1)th day: The user fills in the response load, that is, the expected response load ΔP on the Tth day committed by the user 填报响应 ;

[0087] On the Tth day: Process load reduction, and obtain the normal load on the demand response day based on the prediction model

[0088] On the (T + 1)th day: Obtain the actual load on the demand response day (the Tth day)

[0089] Step S42: Based on the load data in Step S41, calculate the actual load response of the user on the demand response day;

[0090]

[0091] Step S43: Calculate the ratio of the actual load response on the user's demand response day to the expected response amount filled in the previous day, and take it as the user's compliance response degree e.

[0092]

[0093] Step S5: Calculate the compliance response degrees of previous times based on the user's historical demand response data, and obtain the compliance response degrees of the same user under different time periods, different electricity price differences, and different date application scenarios;

[0094] Different time periods: The daily peak electricity consumption periods are from 10:00 to 12:00 in the morning and from 15:00 to 21:00 in the afternoon;

[0095] Different electricity price differences: The demand response during peak load periods is subsidized at a standard of 2 yuan per kilowatt-hour, and the demand response during valley load periods is subsidized at a standard of 0.35 yuan per kilowatt-hour;

[0096] Different dates: Weekdays, weekends, and holidays.

[0097] Taking an industrial user A as an example, calculate the compliance response degrees under its different application scenarios, and the results are shown in Tables 2, 3, and 4.

[0098] Table 2 Compliance response degree of user A on weekdays

[0099]

[0100]

[0101] Table 3 Compliance response degree of user A on weekends

[0102]

[0103] Table 4 User A's compliance response rate during holidays

[0104]

[0105] The present invention evaluates the demand response compliance ability of users based on specific scenarios, that is, based on the daily predicted load obtained from the short-term user load prediction model and the actual load on the day of demand response, calculates the actual load response on the demand response day, and then calculates the ratio with the reported load response before the demand response day, calculates the compliance response rate in different scenarios of different users, so as to accurately evaluate the demand response compliance of users.

[0106] Step S6: Since the number of users participating in demand response is relatively small and it is impossible to accurately evaluate the compliance response rate based on the current data by constructing a complex model, the minimum value of the user demand response compliance response rate in various application scenarios is combined with the user demand response report volume to calculate the evaluation value of the adjustable resource potential on the user side. Based on the evaluation of the compliance ability of user demand response, the demand side formulates a high-potential and compliance white list, and gives priority to the users on the list when conducting demand response invitations, so as to improve the accuracy of grid regulation on the user side and ensure the balance of power grid supply and demand.

[0107] In order to verify the effectiveness of the method proposed in the present invention for evaluating the user compliance response ability, a comparison is made with the existing methods for evaluating the user demand response compliance ability, and the comparison methods are as follows:

[0108] (1) Intelligent business chain: The intelligent business chain is used to evaluate the user compliance response rate;

[0109] (2) Bayesian probability: Bayesian probability is used to evaluate the user compliance response rate, and the compliance evaluation is updated according to the phased compliance behavior;

[0110] (3) LSTM + attention mechanism: LSTM combined with the attention mechanism is used to evaluate the user compliance response rate;

[0111] (4) The method of the present invention.

[0112] To demonstrate the generalization applicability of the method described in the present invention, the electricity load data of industrial and commercial users in 3 different regions are collected, and 3 groups of load sample data sets are constructed as follows:

[0113] Sample group 1: It contains 528 sample data from region A;

[0114] Sample group 2: It contains 319 sample data from region B;

[0115] Three groups of samples: including 96 sample data from Region C;

[0116] The comparison results of the effectiveness of the evaluation method proposed in the present invention are shown in Table 5.

[0117] Table 5 Comparison table of the mean values of the compliance response accuracy

[0118]

[0119]

[0120] As can be seen from Table 5, on the three groups of sample data, the method of the present invention shows good evaluation performance and can effectively evaluate the compliance response degree in the process of user demand response. Through evaluation, the power grid operator can formulate a more reasonable power dispatching plan according to the actual response ability of users, balance the supply and demand relationship of the power system, improve the utilization rate of power resources, contribute to the optimal allocation of power resources, enhance the stability and reliability of the power system, and provide technical support for the refined management of power resources.

[0121] To verify the execution efficiency of the method of the present invention, the time efficiency of the above 4 methods was compared, namely the method of the present invention (Method 1), the intelligent service chain (Method 2), Bayesian probability (Method 3), and LSTM + attention mechanism (Method 4), and the comparison results are as Figure 6 shown.

[0122] From Figure 6 it can be seen that the evaluation method based on the intelligent service chain has the lowest execution time because the intelligent service chain method has less data calculation volume and the mean accuracy rate is about 67.59%; the method of the present invention has the second lowest execution time after the intelligent service chain method, but the highest accuracy rate, with a mean of about 94.66%. At the same time, the execution time is lower than that of Method 3 and Method 4 representing deep learning. This is because the present invention uses a temporal convolutional network to implement parallel computing and processing at each time step, uses the attention mechanism to extract important feature information and performs information selection and fusion through a gating mechanism, which can greatly shorten the information propagation path, better capture long-term dependence relationships, and effectively improve the execution efficiency.

[0123] Embodiment 2

[0124] This embodiment provides a user adjustable load evaluation system for demand response,

[0125] A user adjustable load evaluation system for demand response, comprising:

[0126] A prediction module, configured to: based on the acquired user load data before the demand response day, the meteorological data on the demand response day, and the holiday data, use the trained short-term user power load prediction model to obtain the predicted user power load on the demand response day;

[0127] A performance response degree calculation module, configured to: calculate the actual load response amount on the demand response day according to the difference between the actual user power load amount and the predicted user power load on the demand response day; calculate the performance response degree of the user according to the ratio between the actual load response amount and the reported expected response amount on the demand response day; calculate the performance response degrees of previous times based on the user's historical demand response data, and obtain the performance response degrees of the same user under different time periods, different electricity price differences, and different application scenarios on different dates;

[0128] An evaluation module, configured to: based on the minimum value of the user demand response performance response degree matching the application scenario, combined with the user demand response reported amount, evaluate the potential of the adjustable resources on the user side.

[0129] In some embodiments, the prediction module is specifically configured to: convert the acquired user load data before the demand response day, the meteorological data on the demand response day, and the holiday data into two-dimensional data, respectively model the short-term intra-day time series and the long-term daily periodic pattern, and perform normalization processing; based on the two-dimensional data, use a dual-branch structure to capture information, capture long-term time series features through the long-term time series feature extraction branch, and capture short-term intra-day time series features through the short-term intra-day time series feature extraction branch; fuse the long-term time series features and the short-term intra-day time series features, and through linear processing, obtain the predicted user power load on the demand response day.

[0130] In some embodiments, the long-term time series feature extraction branch adopts a network composed of TCN, GRU, and an attention mechanism, and the short-term intra-day time series feature extraction branch adopts a linear layer; use TCN to extract the long-term historical information features of each time step, and introduce an attention mechanism to automatically screen the important information features in the long-term historical information features of each time step, and calculate the weights of the important information features; input the important information features into GRU for parallel information selection and fusion to obtain long-term time series features. Use a linear layer to aggregate the short-term intra-day fluctuation data to obtain short-term intra-day time series features.

[0131] In some embodiments, the evaluation module is specifically configured to: based on the minimum value of the user demand response performance response degree matching various application scenarios, combined with the user demand response reported amount, calculate the evaluation value of the potential of the adjustable resources on the user side, and based on the evaluation of the performance ability of the user demand response, the demand side formulates a high-potential and performance white list, and gives priority to the users on the list when conducting demand response invitations.

[0132] Embodiment III

[0133] This embodiment provides a computer-readable storage medium, on which a computer program is stored. When the program is executed by a processor, the steps in the demand-response-oriented user-adjustable load evaluation method described in the first embodiment above are implemented.

[0134] Embodiment Four

[0135] This embodiment provides a computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the steps in the demand-response-oriented user-adjustable load evaluation method described in the first embodiment above are implemented.

[0136] Embodiment Five

[0137] This embodiment provides a computer program product or a computer program. The computer program product or the computer program includes computer instructions, and the computer instructions are stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the steps in the demand-response-oriented user-adjustable load evaluation method described in the first embodiment above.

[0138] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a hardware embodiment, a software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memories and optical memories, etc.) containing computer-usable program codes.

[0139] The present invention is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, as well as the combination of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for implementing the functions specified in Figure 1 one or more flows or multiple flows and / or blocks Figure 1 one or more blocks or multiple blocks.

[0140] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a particular manner, such that the instructions stored in the computer-readable memory produce a manufacture including an instruction device that implements the functions specified in one or more processes and / or blocks Figure 1 in one or more processes and / or blocks Figure 1 specified in the block or blocks.

[0141] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operational steps are performed on the computer or other programmable device to produce a computer-implemented process, thereby providing steps for implementing the functions specified in one or more processes and / or blocks Figure 1 in one or more processes and / or blocks Figure 1 specified in the block or blocks.

[0142] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above methods. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM), a random access memory (RAM), etc.

[0143] The foregoing is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention can have various modifications and changes. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A user-adjustable load assessment method for demand response, characterized in that Including: Based on the obtained user load data before the demand response day, the meteorological data on the demand response day, and the holiday data, using the trained user short-term electricity load prediction model to obtain the predicted user electricity load on the demand response day; According to the difference between the actual electricity load of the user on the demand response day and the predicted user electricity load, calculate the actual load response on the demand response day; According to the ratio between the actual load response on the demand response day and the reported expected response, calculate the compliance response degree of the user; Based on the user's historical demand response data, calculate the compliance response degree for each time, and obtain the compliance response degree of the same user under different time periods, different electricity price differences, and different date application scenarios; Based on the minimum value of the user demand response compliance response degree matching the application scenario, combined with the user demand response reported quantity, evaluate the potential of adjustable resources on the user side.

2. The method for evaluating user-adjustable loads for demand response according to claim 1, wherein The process of using the user short-term electricity load prediction model includes: Converting the obtained user load data before the demand response day, the meteorological data on the demand response day, and the holiday data into two-dimensional data, respectively modeling the short-term intra-day time series and the long-term daily periodic pattern, and performing normalization processing; Based on the two-dimensional data, using a double-branch structure for information capture, capturing long-term time series features through the long-term time series feature extraction branch, and capturing short-term intra-day time series features through the short-term intra-day time series feature extraction branch; Fusing the long-term time series features and the short-term intra-day time series features, and through linear processing, obtaining the predicted user electricity load on the demand response day.

3. The method for evaluating user-adjustable load for demand response according to claim 2, wherein, The long-term time series feature extraction branch adopts a network composed of TCN, GRU, and an attention mechanism, and the short-term intra-day time series feature extraction branch adopts a linear layer.

4. The method for evaluating user-adjustable load for demand response according to claim 3, wherein The method for capturing long-term time series features through the long-term time series feature extraction branch includes: using TCN to extract the long-term historical information features of each time step, introducing an attention mechanism to automatically screen the important information features in the long-term historical information features of each time step, and calculating the weights of the important information features; inputting the important information features into GRU for parallel information selection and fusion to obtain long-term time series features.

5. The method for evaluating user-adjustable load for demand response according to claim 3, wherein The method for capturing short-term intra-day time series features through the short-term intra-day time series feature extraction branch includes: using a linear layer to aggregate the short-term intra-day fluctuation data to obtain short-term intra-day time series features.

6. The method for evaluating user-adjustable load for demand response according to claim 1, wherein The method for evaluating the potential of adjustable resources on the user side based on the minimum value of the user demand response compliance response degree matching the application scenario, combined with the user demand response reported quantity includes: based on various application scenarios, matching the minimum value of the user demand response compliance response degree, combined with the user demand response reported quantity, calculating the evaluation value of the potential of adjustable resources on the user side, based on the evaluation of the compliance ability of the user demand response, formulating a high-potential and compliance white list on the demand side, and giving priority to the users on the list when making demand response invitations.

7. A user-adjustable load evaluation system for demand response, characterized in that, Including: A prediction module configured to: based on the obtained user load data before the demand response day, the meteorological data on the demand response day, and the holiday data, using the trained user short-term electricity load prediction model to obtain the predicted user electricity load on the demand response day; A performance response degree calculation module, which is configured to: calculate the actual load response amount on the demand response day according to the difference between the actual user electricity consumption load amount and the predicted user electricity consumption load amount on the demand response day; calculate the performance response degree of the user according to the ratio between the actual load response amount on the demand response day and the reported expected response amount; calculate the performance response degrees of previous times based on the user's historical demand response data, and obtain the performance response degrees of the same user under application scenarios of different time periods, different electricity price differences, and different dates; An evaluation module, which is configured to: evaluate the potential of adjustable resources on the user side based on the minimum value of the performance response degree of the user's demand response matching the application scenario and in combination with the reported amount of the user's demand response.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by a processor, it implements the steps in the method for evaluating user adjustable load for demand response according to any one of claims 1-6.

9. A computer device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps in the method for evaluating user adjustable load for demand response according to any one of claims 1-6.

10. A computer program product, characterized in that, The computer program product includes a computer program, and when the computer program is executed by a processor, it implements the steps in the method for evaluating user adjustable load for demand response according to any one of claims 1-6.

Images