A demand response excitation method applicable to sub-load guidelines

Abstract

This invention discloses a demand response incentive method applicable to sub-load profiles, spanning three levels: Independent System Operator (ISO), Load Aggregator, and Registered Users. The ISO formulates the Load Response Model (CDL) and publishes it along with a first incentive signal to the aggregator. The aggregator decomposes the load profile into multiple Sub-Load Models (SCDLs) based on users' historical electricity consumption behavior and publishes the SCDLs and a second incentive signal to users. The ISO, aggregator, and users, at each level, establish revenue models based on the interactions between different levels. Solving the revenue models of each entity yields optimized first and second incentive signals. This invention can provide guidance for users with different electricity consumption characteristics, thereby reducing the adjustment costs for adjustable users and fully releasing resource flexibility. Furthermore, by establishing revenue models based on the interactions between different levels, the various entities ultimately reach a benefit equilibrium through game theory.

Description

Technical Field

[0001] This invention belongs to the field of electrical engineering and automation technology, and specifically relates to a demand response excitation method applicable to subload guidelines. Background Technology

[0002] To achieve the "net-zero emissions" goal of the Paris Agreement, renewable energy is increasingly replacing fossil fuels and developing rapidly, with its share in electricity production continuously increasing. Therefore, China has proposed a new power system with greater capacity to accommodate renewable energy. However, the uncertainty of renewable energy limits regulation capabilities, resulting in significant supply-side inadequacy and posing a major challenge to the balance of the new power system.

[0003] Insufficient supply-side regulation capacity necessitates seeking solutions from the demand side. In reality, the demand side comprises a large number of users who can modify their electricity consumption behavior according to the grid's objectives. Demand response (DR) seeks regulation capabilities from flexible loads, providing regulation services to the power system and maintaining power balance between supply and demand. To date, large-scale DR trials have been conducted in many regions of China, demonstrating that DR is a promising method for solving the aforementioned problems. To further promote the widespread adoption of disaster recovery, scholars have proposed that disaster recovery in new power systems should possess four characteristics: scalability, standardization, lightweight implementation, and user participation. However, the promotion of DR faces several challenges: Firstly, the large user base, small individual loads, and diverse electricity consumption behaviors mean that regulation requirements often conflict with users' energy consumption patterns. In other words, users must pay significant regulation costs, such as losses in electricity comfort, which impacts their daily lives. Secondly, DR objectives are too singular, with only a small number of users meeting the regulation requirements. Most users choose not to participate because the regulation objectives are inconsistent with their own electricity consumption behaviors. It is difficult to provide personalized guidance to different users. Therefore, it is urgent to design personalized disaster recovery solutions based on user energy consumption behavior and establish an interactive disaster recovery mechanism between users and the power grid.

[0004] Existing technologies attempt to address these issues from different perspectives. Currently widely used pricing schemes can be divided into price-based pricing (PDR) and incentive-based pricing (IDR). For PDR, package pricing, as an innovative approach to electricity market reform, may become a future trend. For example, some literature identifies user behavior and releases customized electricity price packages to optimize user electricity consumption curves. Other literature designs multiple retail price packages based on customer preferences in a two-tier market to maximize revenue. However, these scholars focus on optimizing electricity price packages while neglecting the user's autonomous choice and interaction process. Furthermore, traditional single-objective optimization struggles to resolve the conflict of interest between sales companies and users. To address this issue, game theory offers a new approach. Some literature establishes electricity price game models between electricity sales companies and users to increase their revenue and reduce peak-valley differences. However, package pricing is difficult to implement widely. First, China's electricity market does not allow for arbitrary price adjustments. Multiple pricing mechanisms would place enormous pressure on market regulation. Furthermore, current research on package pricing primarily discusses the power dynamics between electricity retailers and users, without analyzing the impact of different users on their choice of pricing packages, or considering whether adjusting pricing packages can guide users, promote interaction with the grid, and achieve a win-win situation. To achieve IDR (Independent Power Response), a user baseline (CBL) was established to represent each user's power consumption behavior. Then, the level of user participation was assessed by comparing the user's power curve with the CBL. Many scholars have debated the fairness and effectiveness of the CBL, pointing out its limitations and opposing its widespread adoption.

[0005] In recent years, scholars have proposed a demand response mechanism based on the User Load Guideline (CDL), which balances the "adjustable resources" on the demand side with the "unadjustable resources" on the grid side. The CDL defines the required load profile, guides user participation in demand response (DR), and simplifies support for the construction of new power systems. However, current methods have not effectively reduced user regulation costs, nor have they fully stimulated user responses. Scholars have proposed Sub-User Directrix Load (SCDL) based on user consumption behavior, providing a new solution for reducing DR regulation costs. Nevertheless, the lack of a complete and feasible demand response framework and incentive mechanism hinders the practical application of SCDL-based demand response. Developing a DR mechanism and incentive methods suitable for SCDL is crucial to addressing these challenges. Summary of the Invention

[0006] This invention provides a demand response incentive method applicable to sub-load baselines, which aims to achieve a win-win situation at each level through coordination and game theory among entities by spanning three levels: demand response center, load aggregator, and registered user.

[0007] To achieve the above technical objectives, the present invention adopts the following technical solution:

[0008] A demand response excitation method applicable to subload guidelines includes:

[0009] Independent system operators (ISOs) develop a projected load curve, or load baseline, and release the load baseline and a first incentive signal to load aggregators. The first incentive signal refers to the purchase fee paid by the ISO to the load aggregators for demand response, which is used to incentivize each load aggregator to submit optimized power consumption data.

[0010] The load aggregator decomposes the load profile into multiple sub-load profiles based on the user's historical electricity consumption behavior, and publishes the sub-load profiles and a second incentive signal to the user. The second incentive signal refers to the purchase fee paid by the load aggregator to the user for demand response, which is used to incentivize the user to select the sub-load profile as a guiding target to adjust the electricity consumption pattern, thereby obtaining optimized power consumption data for various users.

[0011] Based on the load baseline, the first incentive signal, and the power consumption data submitted by each load aggregator, a revenue model for independent system operators is constructed with the goal of minimizing the operating costs of independent system operators. Among them, the power consumption data submitted by the load aggregator is the total power consumption data predicted by the load aggregator after all users participate in the demand response incentive.

[0012] Based on the load baseline, the first incentive signal, the sub-load baseline, the second incentive signal, and the power consumption data optimized by the user, a revenue model for the load aggregator is constructed with the goal of maximizing the revenue of the load aggregator. Among them, the power consumption data uploaded by the user is the predicted power consumption data after each user participates in the demand response incentive.

[0013] Based on the second incentive signal and the user's own power data, a user-side benefit model is constructed with the goal of maximizing the benefits gained from participating in demand response and the utility of electricity consumption.

[0014] Solve the revenue models for each entity—independent system operator, load aggregator, and user—to obtain optimized first and second incentive signals.

[0015] Furthermore, the load aggregator decomposes the load profile into multiple sub-load profiles based on users' historical electricity consumption behavior, specifically:

[0016] Historical load curves of several users are obtained and normalized. Then, the normalized load curves are clustered to obtain various typical electricity consumption patterns.

[0017] Construct an objective function to minimize the demand response cost of adjustable load under each electricity consumption pattern:

[0018]

[0019] In the formula, It is the load at time t in the load curve corresponding to the i-th electricity consumption mode. This represents the load value at time t within the i-th sub-load guideline; Let W be the set of time points in the load curve, and N be the set of various typical electricity consumption patterns; i The percentage of electricity consumption for users under the i-th electricity consumption mode;

[0020] Establish constraints:

[0021]

[0022] In the formula, P t D* It is the load baseline value at time t published by the independent system operator.

[0023] Furthermore, a revenue model for independent system operators is constructed, specifically as follows:

[0024]

[0025]

[0026] In the formula, C power-grid Represents the operating costs of an independent system operator. The first incentive signal provided by the independent system operator to aggregator j, I j It is the demand response service provided by the independent system operator to aggregator j; M is the set of aggregators; Let d be the power consumption of aggregator j at time t; Euclidean distance d j The similarity between the power curve and the load baseline of aggregator j is used to describe the degree of similarity; ε represents the correction coefficient for the similarity. G The regulation costs paid by independent system operators to conventional adjustable generating units; d j,t The power consumption of aggregator j at time t With load guideline P t D* The Euclidean distance between them, a g b g and c g This represents the cost coefficient.

[0027] Furthermore, the revenue model for the load aggregator is constructed as follows:

[0028]

[0029] In the formula, Represents the revenue of aggregator j. The first incentive signal provided by the independent system operator to aggregator j, I j It is a demand response service provided by an independent system operator to aggregator J; I represents the second incentive signal provided by the aggregator to the k-th type of user. k It is the demand response service provided by the aggregator to the k-th type of user, where K is the set of user types; Let d represent the power consumption of the k-th user at time t. k Power curve used to describe the k-th type of user With sub-load guideline P k D* The degree of similarity between them, where ε represents the similarity correction coefficient. P represents the subload profile of the k-th user class. k D* The power consumption at time t.

[0030] Furthermore, a user-side revenue model is constructed, specifically as follows:

[0031]

[0032]

[0033] In the formula, This represents the revenue of the k-th type of user. I represents the second incentive signal provided by the aggregator to the k-th type of user. k It is the demand response service provided by the aggregator to the k-th type of user. This represents the power consumption of the k-th user at time t. This represents the utility that the k-th type of user derives from electricity consumption at time t. This indicates the maximum load power consumption capability on the user side. This indicates that the k-th type of user is in the time period Total power consumption within, a k and b k For user-dependent parameters.

[0034] Furthermore, the types of load aggregators include industrial, commercial, and residential.

[0035] Beneficial effects

[0036] This invention considers three independent entities at different levels: the ISO (Independent Controller), load aggregators, and demand response (DR) users. First, the ISO issues the Conditional Demand Response (CDL) based on the power system's supply and demand balance, serving as the overall system's regulation demand. The CDL acts as a regulation target, guiding different load aggregators. Second, load aggregators decompose the CDL into different Specific Controller Demand Response (SCDL) based on the different characteristics of their users, providing guidance for users with varying electricity consumption characteristics, thereby reducing the regulation costs for adjustable users and fully releasing resource flexibility. Third, each level incurs regulation costs and earns monetary incentives during demand response implementation. This invention establishes a revenue model based on the interactions between different levels, ultimately achieving a profit equilibrium through game theory. Attached Figure Description

[0037] Figure 1 This is the demand response mechanism architecture applicable to sub-load guidelines as described in this invention;

[0038] Figure 2 This is the demand response implementation process for sub-load baselines as described in the embodiments of the present invention;

[0039] Figure 3 This is a schematic diagram illustrating the interest analysis of different entities as described in the embodiments of the present invention;

[0040] Figure 4 This is the response effect of the power grid in Case 1 of the present invention publishing a unified CDL to users of four different electricity consumption modes. (a)(b)(c)(d) correspond to the four different electricity consumption modes respectively.

[0041] Figure 5 These are the SCDL response results of the aggregator to the DR user agent for four different power consumption modes in Case 2 of the present invention. (a)(b)(c)(d) correspond to the four different power consumption modes respectively.

[0042] Figure 6 These are the SCDL response results of the aggregator for the DR user agent for four different power consumption modes in Case 3 of the present invention, where (a), (b), (c), and (d) correspond to the four different power consumption modes respectively.

[0043] Figure 7 This refers to the adjustment costs of the three load aggregators in different cases by ISO.

[0044] Figure 8 This refers to the revenue of the load aggregator in different cases;

[0045] Figure 9 These are the benefits to users in responding to their needs in different cases;

[0046] Figure 10 It refers to the demand response adjustment costs for users in different cases. Detailed Implementation

[0047] The embodiments of the present invention will be described in detail below. These embodiments are based on the technical solutions of the present invention and provide detailed implementation methods and specific operation processes to further explain the technical solutions of the present invention.

[0048] This invention provides a demand response incentive method applicable to subload baselines, spanning three levels: Independent System Operator (ISO), Load Aggregator, and Registered Users. It aims to achieve a win-win situation for each level through coordination and negotiation among these entities. Specifically, the ISO issues a Conditional Demand Response (CDL) based on the power system's supply and demand balance, serving as the overall system's adjustment demand. The CDL, as the adjustment target, provides guidance for different load aggregators. Load aggregators decompose the CDL into different Subload Response (SCDL) based on the different characteristics of their users. The SCDL provides guidance for users with different electricity consumption characteristics, aiming to minimize the adjustment costs for adjustable users and thus fully release resource flexibility. Each level incurs adjustment costs and earns monetary incentives during the DR implementation process, and the various entities at different levels ultimately reach a state of interest equilibrium through negotiation.

[0049] 1. Independent System Operators (ISOs) develop and publish load baselines and excitation signals.

[0050] In recent years, some scholars have proposed using "adjustable resources" on the demand side to balance the adjustment demand from unadjustable resources, and calculating the expected load curve of adjustable users, namely the load line (CDL). As the adjustment target of demand response users, the CDL can measure the contribution of "adjustable resources" to the supply and demand balance.

[0051] 2. ISO Revenue Model

[0052] After calculating the load baseline based on the supply and demand balance, the grid entity (belonging to an independent system operator) purchases demand response (DR) services from an aggregator (LA) entity to achieve demand response power balancing. It's important to note that if the aggregator cannot provide complete DR regulation services, meaning the aggregated load curve still deviates from the load baseline, the grid center will smooth out the remaining deviation by dispatching adjustable generating units. Since dispatching adjustable units to smooth out non-adjustable fluctuations is costly, the grid center aims to achieve its regulation goals by fully utilizing the aggregator's flexibility resources.

[0053] Therefore, the goal of the power grid center is to minimize operating costs, including the generation costs of conventional adjustable generating units and the costs of purchasing DR services from load aggregators. As shown in equation (6):

[0054]

[0055]

[0056] In the formula, C power-grid The first term of equation (6) represents the operating cost of an independent system operator; the second term C represents the cost of the power grid center purchasing DR services from the aggregator. G The regulation cost payment made by the power grid center to the conventional adjustable generating units is in the form widely adopted by formula (10);

[0057] This represents the first incentive signal provided by the independent system operator to aggregator j; I j It is the demand response service provided by the independent system operator to aggregator j, which can be obtained through equation (7); M is the set of aggregators; Let d be the power consumption of aggregator j at time t; Euclidean distance d j This describes the similarity between the power curve and the load baseline of aggregator j; ε represents the similarity correction coefficient, typically taken as 500-1000; d j,t The power consumption of aggregator j at time t With load guideline P t D* The Euclidean distance between them, a g b g and c g This represents the cost coefficient.

[0058] 3. Load Aggregator (LA) Decomposition Calculation of Sub-Load Guidelines

[0059] Load aggregators with flexible resources play a dual role: they are both providers of demand response (DR) services to ISO (Independent Regulation Authority) and purchasers of DR services for adjustable users. Load aggregators can be understood as large DR users, encompassing numerous flexible loads such as industrial and commercial users. They can optimize the overall load power curve by scheduling their abundant flexible adjustable resources within their footprint, based on the regulation targets issued by ISO. Once they receive the guidance and incentive signals issued by ISO, aggregators first assess the response potential of their adjustable loads, then decompose the demand response management (CDL) based on the different electricity consumption characteristics of each adjustable load to formulate the demand response management (SCDL) that minimizes the adjustable load's demand response adjustment costs. The SCDL, as a regulation signal for adjustable loads, is issued to the user group to guide the flexible users within its jurisdiction to adjust their electricity consumption behavior in response. Furthermore, load aggregators pay for DR services to user groups based on their proximity to the sub-benchmark, which is the incentive signal issued to the user group. The revenue of load aggregators mainly includes DR incentives received from ISO and DR purchase costs paid to user groups.

[0060] The CDL (Conditional Demand Response) defined by ISO does not consider the electricity consumption characteristics of users. In actual demand response implementation, due to the different electricity consumption characteristics of many users, the consumption patterns on the demand side are also different. Therefore, the deviation between the user load curve and the CDL leads to significant demand response adjustment costs. When all users' load curves are perfectly matched with the CDL, the variation from renewable energy (RE) and inelastic loads is minimized. The electrical energy between the user load curve and the CDL, representing the electrical energy required to approach the CDL, is defined as the adjustment cost. This cost is quantified by calculating the norm of the difference between the vectors of the two curves; due to the increasing marginal cost, the 2-norm is used. The mathematical definition of the user's demand response adjustment cost is shown in Equation (1).

[0061]

[0062] Among them, P t D and L t These represent the power values ​​of the load baseline CDL and the user load curve at time t, respectively.

[0063] When a unified load curve is used as the target response, the response results of large-scale adjustable loads may deviate significantly from the unified load curve. Only a small number of users can achieve the system's adjustment target by slightly adjusting the load curve, while the vast majority choose not to participate in the response due to the excessive adjustment cost. In order to fully mobilize users' response enthusiasm, the demand response center has to issue higher incentives to loads with weak enthusiasm, which greatly weakens the system's economy while failing to guarantee the response effect.

[0064] Therefore, due to differences in demand-side electricity consumption characteristics, a unified load profile cannot individually guide adjustable loads to participate in demand response. The load aggregator (LA) of this invention is based on the decomposition of the CDL (Constant Load Level) from the user's historical load curve to obtain several sub-load profiles (SCDL), each SCDL representing a typical electricity consumption pattern.

[0065] First, define the shape of the load curve: Let l t For a load curve, in the time period The total amount of electricity in the container is written as equation (2).

[0066]

[0067] Based on the total electricity α, l t Normalization yields That is, l t After standardization algorithm f u (l t The value after ) is written as equation (3).

[0068]

[0069] By using a load consumption characteristic clustering method, N types of electricity consumption patterns are identified from the standardized historical load curves. The typical load curve for the i-th electricity consumption pattern is then determined. in, This is the unit value after removing dimensions, retaining only the curve shape characteristics of the electricity consumption mode. The proportion of user electricity consumption under the i-th electricity consumption mode is W. i , representing the electricity consumption weight under the i-th electricity consumption mode.

[0070] After obtaining typical load curves under various typical electricity consumption patterns through clustering, the sub-load guidelines under each typical electricity consumption pattern are solved by constructing an objective function and establishing constraints:

[0071] (1) Objective Function: The decomposition objective aims to minimize the demand response cost of adjustable loads under each type of electricity consumption mode. That is, the weighted sum of the squared Euclidean distance between the typical curve under each type of electricity consumption mode and the decomposed SCDL is minimized, written as Equation (4). This is equivalent to classifying user curves with the same electricity consumption mode into a category, in which the user curve can achieve the minimum adjustment cost by selecting the corresponding SCDL. Under the guidance of the SCDL, the shape of the total load curve after the response will be much closer to the CDL.

[0072]

[0073] In the formula, It is the load at time t in the typical load curve corresponding to the i-th electricity consumption mode; W represents the set of time points in a typical load curve, where N is the set of various electricity consumption patterns; i The percentage of electricity consumption for users under the i-th electricity consumption mode; This represents the load value at time t in the i-th sub-load guideline, which is sent from the power grid entity to the user entities under its jurisdiction, responding to the rational decision-making of users on that day to select the SCDL that conforms to their own electricity consumption pattern as the guiding target.

[0074] (2) Constraints: The constraints are shown in equation (5). During the decomposition process, the SCDL value at any time t is guaranteed to be consistent with the power weight W of the corresponding power consumption mode category. i The sum of the products equals the CDL value at time t. In terms of guidance, SCDL is equivalent to CDL, with the value at each time t between [0,1], and the cumulative sum is 1.

[0075]

[0076] In the formula, P t D* It is the load baseline value at time t published by the independent system operator.

[0077] 4. Revenue Model of Load Aggregator

[0078] For aggregator entities, on the one hand, they should maximize the incentives they receive from the grid, which is determined by the similarity between the aggregator's total power consumption curve and the load curve. On the other hand, aggregators should minimize the incentive costs issued to DR user entities. Therefore, the aggregator's revenue function is shown in equation (11).

[0079]

[0080] The first term of Equation (11) represents the cost of purchasing DR services from the aggregator by the power grid center, and the second term represents the incentive cost issued by the aggregator i to its user groups.

[0081] In the formula, Represents the revenue of aggregator j. The first incentive signal provided by the independent system operator to aggregator j, I j It is a demand response service provided by an independent system operator to aggregator J; I represents the second incentive signal provided by the aggregator to the k-th type of user. k It is the demand response service provided by the aggregator to the k-th type of user, where K is the set of user types; Let d represent the power consumption of the k-th user at time t. k Power curve used to describe the k-th type of user With sub-load guideline P k D* The degree of similarity between them P represents the subload profile of the k-th user class. k D* The power consumption at time t.

[0082] 5. User and Revenue Model

[0083] User aggregators with flexible resources provide regulation (DR) services. Virtual power plants, industrial and commercial users, and other user-level entities can be certified as DR users through registration, while unregistered users are considered unregisterable loads. After receiving the SCDL (Standardized Regulation Principles) issued by the aggregator, users rationally choose the sub-benchmark closest to their own electricity consumption behavior as the regulation target to participate in the response, thereby reducing regulation costs. In terms of benefits, user groups mainly benefit from the incentive signals issued by the aggregator and the electricity benefits generated from their own energy consumption.

[0084] The goal of user decision-making is to maximize their electricity benefits, including the benefits gained from participating in DR and the utility of electricity consumption under corresponding constraints, as shown below:

[0085]

[0086] In the formula, This represents the revenue of the k-th type of user. I represents the second incentive signal provided by the aggregator to the k-th type of user. k It is the demand response service provided by the aggregator to the k-th type of user. This represents the power consumption of the k-th user at time t. This represents the utility that the k-th type of user derives from electricity consumption at time t. This indicates the maximum load power consumption capability on the user side. This indicates that the k-th type of user is in the time period Total power consumption within.

[0087] A decrease in the utility of electricity consumption indicates an increase in the loss of user comfort. Without loss of generality, the utility of electricity consumption in this embodiment adopts a piecewise quadratic utility function, as shown in equation (17):

[0088]

[0089] In the formula, a k and b k The user-dependent parameters, which have different values ​​at different time periods, describe the user's time-varying power consumption preferences.

[0090] 6. Analysis of the main interests

[0091] By analyzing the objective functions of the revenue models for each entity, trade-offs between costs and benefits can be found within the objectives of each optimizing entity. For the power grid center, to minimize operating costs—that is, to meet the system's regulation needs as much as possible through scheduling flexibility resources—it needs to improve load regulation performance by increasing incentives for load aggregators, which will lead to higher fees paid to load aggregators. For load aggregators, to maximize their revenue, they need to maximize their revenue from the power grid center while minimizing their DR service fees paid to users. However, increasing DR service fees paid to users can simultaneously increase revenue from the power grid center. Similarly, users also need to weigh their power consumption utility against the revenue they receive from aggregators, because choosing to participate in the response of sub-guidelines will increase their revenue but will also lead to a loss of utility.

[0092] Therefore, the incentives provided by the power grid center affect the regulation performance of users under the jurisdiction of load aggregators, while the incentives provided by load aggregators affect the flexibility provided by users. Conversely, user responses affect load aggregator incentive adjustments, thereby affecting the DR purchase fees and incentive signals paid by the power grid center to aggregators. The interaction between these entities can be summarized as follows: the power grid center issues incentive signals to each load aggregator. Each aggregator submitted its optimized total power consumption vector. In response to receiving the incentive signal, the aggregator simultaneously broadcasts the incentive signal to the user group. The user publishes an optimized self-power vector based on the selected sub-guideline as the guiding target. This is in response to the received excitation signal.

[0093] 7. Response Implementation Process

[0094] This invention relates to a demand response incentive method for sub-load profiles. First, the ISO (Industrial and Environmental Standards Authority) formulates a unified load profile (CDL) based on the renewable energy and non-adjustable load curves in the system. The ISO then distributes the CDL to load aggregators. Aggregators manage numerous types of DR (Demand, Response, and Distribution) user resources. Based on the adjustment capabilities of their managed users, they predict adjustable load curves and formulate SCDLs (Sustainable Load Profiles) based on users' historical electricity consumption information. Aggregators report the predicted adjustable power curves of users to the ISO and publish the SCDLs to the user group. The ISO provides subsidies to aggregators for their reporting performance, and aggregators also provide subsidies based on user response. During this process, the shape of the curves reported by the aggregators and the user response are dynamically changing, and the incentives provided by the ISO to the aggregators and by the aggregators to the user group are also dynamically adjusted, ultimately achieving a three-tiered equilibrium of benefits.

[0095] 8. Experimental Comparison

[0096] Using real-world data from the United States, a city-level test system was developed to validate the proposed demand response (DR) scheme. It included three large load aggregators (LAs) for industrial, commercial, and residential users. User parameters were extended to different time intervals, creating an hourly power consumption sequence for 100 users across 24 hours per day for 31 days in 2022. The demand response deployment based on the SCDL (Sustainable Consumption Daily Limit) was developed using data from the first 30 days, and the deployment was tested using data from day 31. Numerical tests were conducted on a laptop with an Intel Core i7-8565U 1.80GHz CPU and 8GB of RAM, using MATLAB 2021a and gurobi 9.0.

[0097] Based on the established testing system, the effectiveness of the DR mechanism and incentive method based on SCDL was discussed through simulation tests on urban systems. A performance comparison analysis of the proposed method was also conducted, comparing its effectiveness in reducing user demand response adjustment costs and its solutions for balancing the interests of different stakeholders from three perspectives.

[0098] Three examples demonstrate the effectiveness of the proposed demand response incentive method for SCDL. In Example 1, the ISO directly publishes the CDL to the user entity (without the load aggregator specifying the SCDL or incentive signal). In Example 2, the ISO publishes the CDL to the load aggregator entity, which then publishes the SCDL to the user entity without an incentive signal. Example 3 employs the method proposed in this invention.

[0099] 8.1 SCDL Response Results

[0100] Figure 4 This demonstrates the response effect of the power grid in Case 1 publishing a unified CDL to users of four different electricity consumption modes. Figure 5 , Figure 6 The SCDL results for the aggregator's four electricity consumption patterns for DR user agents are presented in Case 2 and Case 3, respectively. In terms of electricity consumption pattern characteristics, the curves for Pattern 1 and Pattern 4 show weaker fluctuations throughout the day, while the curves for Pattern 2 and Pattern 3 show larger fluctuations throughout the day.

[0101] Comparing the user's DR regulation costs under different scenarios reveals a significant gap between load consumption and CDL in Case 1, particularly evident in Modes 2 and 3. This indicates that aligning the user with CDL is costly. Cases 2 and 3 utilize SCDL as a guide from LA1 to the user. Figure 7 and Figure 8 The significant reduction in the shaded area compared to Case 1 demonstrates that using SCDL effectively reduces the user's DR conditioning costs. In Case 3, the user's load profile is closer to SCDL than in Case 2, indicating that the user has further optimized their load profile under the proposed excitation method. This can improve response efficiency in each power mode.

[0102] 8.2 Classification of Electricity Consumption Modes

[0103] The following outlines the benefits across three layers for different entities, including the costs and benefits for ISO, load aggregators, and users. LA1, LA2, and LA3 represent different types of users in the city system. By analyzing users' electricity consumption behavior, the users represented by LA1 are categorized into four electricity consumption patterns, while the users represented by LA2 and LA3 are categorized into two and three historical electricity consumption patterns, respectively.

[0104] (1) ISO Benefit Analysis

[0105] Figure 7This shows the regulation costs paid by ISO to three load aggregators in three scenarios. Comparing Case 1, Case 2, and Case 3, it is clear that ISO's demand response regulation costs are significantly lower in Case 3. This reduction is due to ISO's incentives for load aggregators, which in turn incentivize them to explore and utilize user flexibility. By publishing SCDLs and providing incentives to users, load aggregators can more closely align aggregated load consumption with ISO's overall regulation target CDL. Therefore, ISO needs to spend less on thermal power unit regulation to achieve better DR performance at a lower cost.

[0106] The payments made by ISO to LA1, LA2, and LA3 clearly show that ISO pays less to LA1 and more to LA3. This indicates that the electricity consumption behavior of LA1 users is closest to ISO's regulation objectives. Conversely, LA3 users are least similar to CDL users, resulting in higher regulation costs for ISO.

[0107] (2) Revenue Analysis of Load Aggregator

[0108] Figure 8 The data shows the revenue of different load aggregators (LAs) under three scenarios. LA revenue consists of two parts: the cost paid by the ISO for the DR service and the incentives provided by the LAs to users. Comparing the three scenarios, it is clear that only the LAs generate revenue in Cases 2 and 3, while the revenue is higher in Case 3. This indicates that, using the proposed method, the ISO incentivizes the load aggregator, who then publishes the SCDL and provides incentives to users. This coordination among the three layers balances the interests, which is reflected in the higher revenue for the load aggregator in Case 3.

[0109] (3) User revenue analysis

[0110] Figure 9 The revenue of users represented by different LAs is shown in three scenarios. Comparing the two scenarios, it is clear that users in scenario 3 have higher revenue across all LAs. As shown in Equation (13), this revenue includes the revenue from participating in DR and power efficiency. Users represented by LA1 and LA2 have higher revenue than users represented by LA3, indicating that their electricity consumption behavior is closer to the regulatory target. Conversely, users represented by LA3 have less similar behavior to the target and have lower revenue.

[0111] Figure 10The data shows the DR adjustment costs for users represented by different LAs in three scenarios. Comparing the two scenarios, users in scenario 3 have higher DR adjustment costs across all LAs. Users represented by LA1 and LA2 have lower adjustment costs than those represented by LA3, indicating that their behavior is more aligned with the adjustment objectives. This inverse relationship between DR adjustment costs and income corresponds to the previous analysis.

[0112] The above embodiments are preferred embodiments of this application. Those skilled in the art can make various changes or improvements based on them. Without departing from the overall concept of this application, these changes or improvements should fall within the scope of protection claimed in this application.

Claims

1. A demand response excitation method applicable to sub-load guidelines, characterized in that, include: Independent system operators (ISOs) develop a projected load curve, or load baseline, and release the load baseline and a first incentive signal to load aggregators. The first incentive signal refers to the purchase fee paid by the ISO to the load aggregators for demand response, which is used to incentivize each load aggregator to submit optimized power consumption data. The load aggregator decomposes the load profile into multiple sub-load profiles based on the user's historical electricity consumption behavior, and publishes the sub-load profiles and a second incentive signal to the user. The second incentive signal refers to the purchase fee paid by the load aggregator to the user for demand response, which is used to incentivize the user to select the sub-load profile as a guiding target to adjust the electricity consumption pattern, thereby obtaining optimized power consumption data for various users. Based on the load baseline, the first incentive signal, and the power consumption data submitted by each load aggregator, a revenue model for independent system operators is constructed with the goal of minimizing the operating costs of independent system operators. Among them, the power consumption data submitted by the load aggregator is the total power consumption data predicted by the load aggregator after all users participate in the demand response incentive. Based on the load baseline, the first incentive signal, the sub-load baseline, the second incentive signal, and the power consumption data optimized by the user, a revenue model for the load aggregator is constructed with the goal of maximizing the revenue of the load aggregator. Among them, the power consumption data uploaded by the user is the predicted power consumption data after each user participates in the demand response incentive. Based on the second incentive signal and the user's own power data, a user-side benefit model is constructed with the goal of maximizing the benefits gained from participating in demand response and the utility of electricity consumption. Solve the revenue models for each entity—independent system operator, load aggregator, and user—to obtain optimized first and second incentive signals.

2. The demand response incentive method according to claim 1, characterized in that, The load aggregator decomposes the load profile into multiple sub-load profiles based on users' historical electricity consumption behavior, specifically: Historical load curves of several users are obtained and normalized. Then, the normalized load curves are clustered to obtain various typical electricity consumption patterns. Construct an objective function to minimize the demand response cost of adjustable load under each electricity consumption pattern: In the formula, It is the load at time t in the load curve corresponding to the i-th electricity consumption mode. This represents the load value at time t within the i-th sub-load guideline; Let W be the set of time points in the load curve, and N be the set of various typical electricity consumption patterns; i The percentage of electricity consumption for users under the i-th electricity consumption mode; Establish constraints: In the formula, P t D* It is the load baseline value at time t published by the independent system operator.

3. The demand response incentive method according to claim 1, characterized in that, The revenue model for an independent system operator is constructed as follows: In the formula, C power-grid Represents the operating costs of an independent system operator. The first incentive signal provided by the independent system operator to aggregator j, I j It is the demand response service provided by the independent system operator to aggregator j; M is the set of aggregators; Let d be the power consumption of aggregator j at time t; Euclidean distance d j The similarity between the power curve and the load baseline of aggregator j is used to describe the degree of similarity; ε represents the correction coefficient for the similarity. G The adjustment costs paid on behalf of independent system operators to conventional adjustable generating units; d j,t The power consumption of aggregator j at time t With load guideline P t D* The Euclidean distance between them, a g b g and c g Represents the cost coefficient; This is the set of time points in the load curve.

4. The demand response incentive method according to claim 1, characterized in that, The revenue model for the load aggregator is constructed as follows: In the formula, Represents the revenue of aggregator j. The first incentive signal provided by the independent system operator to aggregator j, I j It is a demand response service provided by an independent system operator to aggregator J; I represents the second incentive signal provided by the aggregator to the k-th type of user. k It is the demand response service provided by the aggregator to the k-th type of user, where K is the set of user types; Let d represent the power consumption of the k-th user at time t. k Power curve used to describe the k-th type of user With sub-load guideline P k D* The degree of similarity between them, where ε represents the similarity correction coefficient. P represents the subload profile of the k-th user class. k D* Power consumption at time t, This is the set of time points in the load curve.

5. The demand response incentive method according to claim 1, characterized in that, Construct a user-side revenue model, specifically as follows: In the formula, This represents the revenue of the k-th type of user. I represents the second incentive signal provided by the aggregator to the k-th type of user. k It is the demand response service provided by the aggregator to the k-th type of user. This represents the power consumption of the k-th user at time t. This represents the utility that the k-th type of user derives from electricity consumption at time t. This indicates the maximum load power consumption capability on the user side. This indicates that the k-th type of user is in the time period Total power consumption within, a k and b k For user-dependent parameters.

6. The demand response incentive method according to claim 1, characterized in that, Load aggregators can be categorized into industrial, commercial, and residential types.

Images