Community-level load aggregator regulation and control method and system based on machine learning

Abstract

The invention relates to a community-level load aggregator regulation and control method and system based on machine learning, and belongs to the technical field of load aggregator regulation and control. The method comprises the steps of obtaining community residential electricity consumption data, obtaining residential electricity consumption load information through a residential electricity consumption load identification model, and obtaining a residential response model through a residential cost objective function and a residential response decision variable according to the residential electricity consumption load information; market electricity price information is obtained, and a load aggregator regulation and control model is obtained through the load aggregator income target function and the load aggregator regulation and control decision variable; a resident electricity price excitation response model is obtained according to the combination of the resident response model and the load aggregator regulation and control model, the day-ahead electricity price of the load aggregator is obtained through double-target optimization solution, electricity price pricing is carried out according to the day-ahead electricity price of the load aggregator, electricity utilization of residents in the off-peak period is excited, meanwhile, the benefits of the load aggregator are optimized, and the electricity utilization efficiency of the residents is improved. Therefore, the load regulation and control capability is enhanced and the power grid regulation performance is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of load aggregator regulation and control, and in particular, relates to a community-level load aggregator regulation and control method and system based on machine learning. Background Art

[0002] The current community-level load regulation faces the following situation: from the power supply side, the scale and proportion of new energy installed capacity continue to increase; from the demand side, Shanghai's electricity load continues to rise, and the peak-to-valley difference in the power grid load is increasing year by year. The demand-side resource regulation in the new power system is very urgent, and it is urgent to achieve resource regulation through grid-load interaction and other means to ensure the flexibility and operational stability of the power system.

[0003] In addition, the current load regulation resources are mainly industrial and commercial users. User load regulation is mainly achieved by adjusting load power, generator output power, voltage and frequency, and power topology, mainly through orderly power consumption, virtual power plants, etc. However, compared with large-scale load regulation individuals such as industrial enterprises, community load regulation based on residents is more flexible. From the practice of foreign countries, the load can be effectively reduced by regulating residents' electricity consumption through demand response. However, in China, the potential of residents' load regulation has not been fully tapped and further research and practice are still needed.

[0004] In view of the insufficient application of existing grid-load interaction in community load aggregator regulation, it is urgent to explore a new community-level load aggregator regulation method to achieve transferable load regulation on the resident side and promote the improvement of the grid's adjustable capacity. Summary of the invention

[0005] In order to solve the above problems existing in the prior art, the present invention proposes a community-level load aggregator control method and system based on machine learning.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] A community-level load aggregator control method based on machine learning, the implementation of the community-level load aggregator control method comprising the following steps:

[0008] Acquire community residents' electricity consumption data, and obtain residents' electricity load information through a residents' electricity load identification model based on the community residents' electricity consumption data;

[0009] According to the residential electricity load information, a resident cost objective function is obtained through resident cost minimization modeling, and a resident response model is obtained according to the resident cost objective function and resident response decision variables;

[0010] Obtaining market electricity price information, obtaining a load aggregator profit objective function through load aggregator profit maximization modeling based on the market electricity price information, and obtaining a load aggregator regulation model based on the load aggregator profit objective function and load aggregator regulation decision variables;

[0011] A resident electricity price incentive response model is obtained by combining the resident response model and the load aggregator regulation model. The load aggregator's day-ahead electricity price is obtained by solving the resident electricity price incentive response model through dual-objective optimization. Load aggregator regulation is achieved through electricity price pricing based on the load aggregator's day-ahead electricity price.

[0012] Preferably, the acquiring of community residents' electricity consumption data and obtaining residents' electricity load information through a residents' electricity load identification model according to the community residents' electricity consumption data include:

[0013] Collecting community residents' electricity consumption data through smart meters and eliminating dimensions through data normalization, wherein the community residents' electricity consumption data includes active power of residents' electricity consumption and effective value of residents' electricity consumption current;

[0014] According to the community residents' electricity consumption data, the load equipment operation objective function is solved by an improved GWO algorithm to obtain the residents' electricity load information.

[0015] Preferably, the step of solving the load equipment operation objective function by improving the GWO algorithm according to the community residents' electricity consumption data to obtain the residents' electricity load information includes:

[0016] Initializing a population, wherein the load equipment operation objective function values ​​of individuals in the population are used as fitness values;

[0017] Update the individual positions of the population through differential perturbations;

[0018] The GWO algorithm is continuously iterated to select the individual with the highest fitness value as the optimal solution, that is, the residents' electricity load information.

[0019] Preferably, the resident cost objective function obtained by resident cost minimization modeling according to the resident electricity load information includes:

[0020] Preset the time interval for power load transfer;

[0021] Preset transferable load power range;

[0022] The mathematical expression of the resident cost objective function is:

[0023]

[0024] C=α(t4-t2)Q

[0025] Among them, M tis the real-time electricity price per unit time the day before, u is the basic load of residential electricity consumption, is the activation variable of device j of resident i at time t, is the j-device power of resident i, the original operation time period of the transferable load is [t1, t2], the operation time period after the transferable load is transferred is [t3, t4], α is the resident comfort cost per unit time, Q is the total transfer power of the transferable load, C is the resident comfort cost, and Fa is the resident cost objective function;

[0026] The constraints of the resident cost objective function are that the transferred load must be within the electricity load transfer time interval, the power after the load transfer must be within the transferable load power interval, and the required electric energy before and after the load transfer remains unchanged. The optimized load curve value is not less than 0, and the minimum load after residents participate in demand response is not lower than the minimum load before participating in demand response.

[0027] Preferably, the load aggregator profit objective function obtained by load aggregator profit maximization modeling according to the market electricity price information includes:

[0028] Preset the real-time electricity price range for the day before;

[0029] The mathematical expression of the load aggregator's revenue objective function is:

[0030] Fb=max{H-cost1-cost2-Cr}

[0031]

[0032] Among them, cost1 is the electricity purchase cost of the load aggregator, cost2 is the equipment loss cost of the load aggregator, Cr is the risk cost of the load aggregator, Fb is the load aggregator's profit objective function, H is the load aggregator's power output profit, M t is the real-time electricity price per unit at the day before, and B is the resident response decision variable;

[0033] The constraints of the load aggregator's revenue objective function are that the real-time electricity price per unit time the day before must be within the real-time electricity price range per unit time the day before, the electricity price the day before is equal to the average electricity price the day before, and the load power remains balanced.

[0034] Preferably, the method of solving the residential electricity price incentive response model through dual-objective optimization to obtain the load aggregator's day-ahead electricity price includes:

[0035] Obtaining resident response decision variable values ​​through a Cplex solver according to the resident response model;

[0036] The load aggregator control model is solved by an improved genetic algorithm according to the resident response decision variable value to obtain the load aggregator's day-ahead electricity price.

[0037] Preferably, solving the load aggregator regulation model by using an improved genetic algorithm according to the resident response decision variable value to obtain the load aggregator's day-ahead electricity price includes:

[0038] The resident response decision variable value is substituted into the load aggregator profit objective function as a known condition;

[0039] Initialize the population through Logistic chaos mapping;

[0040] The load aggregator profit objective function value of the individuals in the population is taken as the fitness value, and the individual with the highest fitness value is selected through iterative selection of the genetic algorithm as the optimal solution, that is, the day-ahead electricity price of the load aggregator.

[0041] A community-level load aggregator control system based on machine learning, the system is applied to the above-mentioned community-level load aggregator control method, including a data acquisition module, a residential power load identification module, a residential cost modeling module, a load aggregator revenue modeling module, a residential electricity price incentive response module, and an electricity price pricing and control module;

[0042] The data acquisition module is used to obtain community residents' electricity consumption data and market electricity price information;

[0043] The residential power load identification module is used to obtain residential power load information based on community residents' power consumption data through a residential power load identification model;

[0044] The resident cost modeling module is used to construct a resident response model based on the community residents' electricity consumption data through the resident cost objective function and the resident response decision variables;

[0045] The load aggregator revenue modeling module is used to construct a load aggregator regulation model based on market electricity price information through a load aggregator revenue objective function and a load aggregator regulation decision variable;

[0046] The residential electricity price incentive response module is used to combine the resident response model and the load aggregator regulation model to form a residential electricity price incentive response model, and determine the load aggregator's day-ahead electricity price through a dual-objective optimization solution to balance the minimization of resident costs and the maximization of load aggregator benefits;

[0047] The electricity price pricing and regulation module is used to set electricity prices according to the optimized day-ahead electricity price of the load aggregator and implement load aggregator regulation through a feedback mechanism to encourage residents to use electricity during off-peak hours and optimize the load curve.

[0048] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned community-level load aggregator control method when executing the program.

[0049] A storage medium containing computer executable instructions, which are used to execute the community-level load aggregator regulation method described above when executed by a computer processor.

[0050] The beneficial effects of the present invention are:

[0051] (1) A resident response model is constructed by minimizing resident costs, and a load aggregator regulation model is constructed by maximizing the benefits of load aggregators. A resident electricity price incentive response model is built by combining the resident response model and the load aggregator regulation model, and the load aggregator's day-ahead electricity price is obtained through dual-objective optimization. Different electricity prices are set in different time periods to encourage residents to actively participate in peak load regulation, promote valley electricity consumption by community users participating in demand response, reduce unnecessary electricity resources during peak hours, optimize the load curve, and improve grid utilization efficiency.

[0052] (2) The residential electricity load information is obtained through the residential electricity load identification model. The diversity of the population is increased through differential perturbations. The generated new solutions can help the GWO algorithm explore different areas of the search space, escape from local optimality, and avoid premature convergence, thereby enhancing the global search capability and obtaining high-quality residential electricity load information, providing accurate load information for subsequent control strategies.

[0053] (3) The resident response decision variable value is obtained through the Cplex solver and substituted into the load aggregator's profit objective function, and then the load aggregator's day-ahead electricity price is obtained through the improved genetic algorithm. The introduction of Logistic chaotic mapping to initialize the population can ensure the good distribution of the initial population in the solution space, so that the genetic algorithm can search a wider area and avoid falling into the local optimum. The globally optimized day-ahead electricity price of the load aggregator can be obtained, thereby balancing the minimization of resident costs and the maximization of load aggregator profits. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings.

[0055] Figure 1 It is a flow chart of a community-level load aggregator control method based on machine learning of the present invention. DETAILED DESCRIPTION

[0056] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation methods, structures, features and effects of the present invention are described in detail below in conjunction with the accompanying drawings and preferred embodiments.

[0057] See also Figure 1, a community-level load aggregator control method based on machine learning, comprising:

[0058] S1: Obtaining electricity consumption data of community residents, and obtaining residential electricity load information through a residential electricity load identification model according to the community residents' electricity consumption data;

[0059] S2: obtaining a resident cost objective function through resident cost minimization modeling according to the resident electricity load information, and obtaining a resident response model according to the resident cost objective function and resident response decision variables;

[0060] S3: Obtain market electricity price information, obtain a load aggregator profit objective function through load aggregator profit maximization modeling according to the market electricity price information, and obtain a load aggregator regulation model according to the load aggregator profit objective function and load aggregator regulation decision variables;

[0061] S4: A residential electricity price incentive response model is obtained based on the combination of the resident response model and the load aggregator regulation model, the load aggregator's day-ahead electricity price is obtained by solving the residential electricity price incentive response model through dual-objective optimization, and load aggregator regulation is achieved through electricity price pricing based on the load aggregator's day-ahead electricity price.

[0062] In this embodiment, the obtaining of community residents' electricity consumption data and obtaining the residents' electricity load information through the residents' electricity consumption load identification model according to the community residents' electricity consumption data are specifically implemented by the following steps:

[0063] S101: Collecting electricity consumption data of community residents through smart meters and eliminating dimensions through data normalization, wherein the electricity consumption data of community residents include active power of residents' electricity and effective value of current of residents' electricity;

[0064] S102: solving the load equipment operation objective function by using an improved GWO algorithm according to the community residents' electricity consumption data to obtain residents' electricity load information;

[0065] S102-1: Initializing a population, wherein the operating objective function value of the load equipment of the individuals in the population is used as the fitness value;

[0066] S102-2: Update the individual positions of the population through differential perturbation;

[0067] Specifically, the mathematical expression of the differential disturbance is:

[0068] CH i，n+1 =L r1，n +w×(L r2，n -L r3，n ) Among them, CH i,n+1 is the solution of the i-th individual in the next generation, L r,nis the vector of the rth individual randomly selected in the current generation, r1, r2, r3 are randomly selected different indexes, and w is the differential perturbation weight factor;

[0069] S102-3: The GWO algorithm is continuously iterated to select the individual with the highest fitness value as the optimal solution, that is, the residential electricity load information.

[0070] The mathematical expression of the load equipment operation objective function is:

[0071]

[0072] Where n is the number of residential electrical equipment, P is the active power characteristic of n equipment, I is the effective current characteristic of n equipment, P(s) is the total power at time s, I(s) is the effective current at time s, a is the weight factor of active power, b is the weight factor of effective current, e i is the operating status of the ith electrical equipment of the resident;

[0073] Specifically, the operating status of residential electrical equipment includes normal operation and standby. In the process of solving the load equipment operation objective function, the operating status of residential electrical equipment is used as the decision variable. The residential electrical equipment is identified by fitting the residential electricity active power and the residential electricity current effective value to obtain the residential electricity load information.

[0074] In this embodiment, the resident cost objective function is obtained by resident cost minimization modeling according to the resident electricity load information through the following steps:

[0075] S201: Preset a time interval for power load transfer;

[0076] S202: Preset a transferable load power range;

[0077] S203: The mathematical expression of the resident cost objective function is:

[0078]

[0079] C=α(t4-t2)Q

[0080] Among them, M t is the real-time electricity price per unit time the day before, u is the basic load of residential electricity consumption, is the activation variable of device j of resident i at time t, is the j-device power of resident i, the original operation time period of the transferable load is [t1, t2], the operation time period after the transferable load is transferred is [t3, t4], α is the resident comfort cost per unit time, Q is the total transfer power of the transferable load, C is the resident comfort cost, and Fa is the resident cost objective function;

[0081] Specifically, the unit time resident comfort cost is the comfort cost of residents affected by the shift unit time after the transferable load is transferred, and the equipment startup variable takes a value of 0 or 1, 0 represents equipment standby, and 1 represents equipment startup;

[0082] S204: The constraints of the resident cost objective function are that the transferred load must be within the electricity load transfer time interval, the power after the load transfer must be within the transferable load power interval, and the required electric energy before and after the load transfer remains unchanged. The optimized load curve value is not less than 0, and the minimum load after the residents participate in the demand response is not lower than the minimum load before participating in the demand response.

[0083] In this embodiment, the resident response decision variable is the electricity consumption at each moment after the resident participates in the demand response.

[0084] In this embodiment, the load aggregator profit objective function is obtained by modeling the load aggregator profit maximization according to the market electricity price information through the following steps:

[0085] S301: Preset the real-time electricity price range for the day before;

[0086] S302: The mathematical expression of the load aggregator's revenue objective function is:

[0087] Fb=max{H-cost1-cost2-Cr}

[0088]

[0089] Among them, cost1 is the electricity purchase cost of the load aggregator, cost2 is the equipment loss cost of the load aggregator, Cr is the risk cost of the load aggregator, Fb is the load aggregator's profit objective function, H is the load aggregator's power output profit, M t is the real-time electricity price per unit at the day before, and B is the resident response decision variable;

[0090] S303: The constraints of the load aggregator's profit objective function are that the real-time electricity price per unit time the day before must be within the real-time electricity price range per unit time the day before, the electricity price the day before is equal to the average electricity price the day before, and the load power remains balanced.

[0091] In this embodiment, the load aggregator's control decision variable is the real-time electricity price at each moment.

[0092] In this embodiment, solving the residential electricity price incentive response model through dual-objective optimization to obtain the load aggregator's day-ahead electricity price is specifically implemented by the following steps:

[0093] S401: Obtaining resident response decision variable values ​​through a Cplex solver according to the resident response model;

[0094] S402: Solving the load aggregator control model by using an improved genetic algorithm according to the resident response decision variable value to obtain the load aggregator's day-ahead electricity price;

[0095] S402-1: Substituting the resident response decision variable value into the load aggregator profit objective function as a known condition;

[0096] S402-2: Initialize the population through Logistic chaos mapping;

[0097] Specifically, the mathematical expression of the Logistic chaotic map is:

[0098] g i+1 =β×g i ×(1-g i )

[0099] Among them, g i+1 is the next value of the chaotic sequence, g i is the current value of the chaotic sequence, β is the control factor of the Logistic chaotic mapping;

[0100] S402-3: The load aggregator profit objective function value of the individuals in the population is used as the fitness value, and the individual with the highest fitness value is selected through iterative genetic algorithm as the optimal solution, that is, the day-ahead electricity price of the load aggregator.

[0101] A community-level load aggregator control system based on machine learning includes a data acquisition module, a residential electricity load identification module, a residential cost modeling module, a load aggregator revenue modeling module, a residential electricity price incentive response module, and an electricity price pricing and control module;

[0102] The data acquisition module is used to obtain community residents' electricity consumption data and market electricity price information;

[0103] The residential power load identification module is used to obtain residential power load information based on community residents' power consumption data through a residential power load identification model;

[0104] The resident cost modeling module is used to construct a resident response model based on the community residents' electricity consumption data through the resident cost objective function and the resident response decision variables;

[0105] The load aggregator revenue modeling module is used to construct a load aggregator regulation model based on market electricity price information through a load aggregator revenue objective function and a load aggregator regulation decision variable;

[0106] The residential electricity price incentive response module is used to combine the resident response model and the load aggregator regulation model to form a residential electricity price incentive response model, and determine the load aggregator's day-ahead electricity price through a dual-objective optimization solution to balance the minimization of resident costs and the maximization of load aggregator benefits;

[0107] The electricity price pricing and regulation module is used to set electricity prices according to the optimized day-ahead electricity price of the load aggregator and implement load aggregator regulation through a feedback mechanism to encourage residents to use electricity during off-peak hours and optimize the load curve.

[0108] The computer storage medium of the embodiment of the present invention may adopt any combination of one or more computer-readable media. The computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples (non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, device or device.

[0109] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0110] The program code included on the computer readable medium can be transmitted with any appropriate medium, including but not limited to wireless, electric wire, optical cable, RF, etc., or any suitable combination of the above. The computer program code for performing the operation of the present invention can be written in one or more programming languages ​​or their combinations, and the programming language includes object-oriented programming languages-such as Java, Smalltalk, C++, and also includes conventional procedural programming languages-such as "C" language or similar programming languages. The program code can be executed completely on the user's computer, partially on the user's computer, as an independent software package, partially on the user's computer and partially on the remote computer, or completely on the remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (for example, using an Internet service provider to connect through the Internet).

[0111] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Although the present invention has been disclosed as a preferred embodiment as above, it is not used to limit the present invention. Any technical personnel in this field can make some changes or modify the technical contents disclosed above into equivalent embodiments without departing from the scope of the technical solution of the present invention. However, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present invention without departing from the content of the technical solution of the present invention still fall within the scope of the technical solution of the present invention.

Claims

1. A community-level load aggregator control method based on machine learning, characterized in that: The implementation of the community-level load aggregator control method includes the following steps: Acquire community residents' electricity consumption data, and obtain residents' electricity load information through a residents' electricity load identification model based on the community residents' electricity consumption data; According to the residential electricity load information, a resident cost objective function is obtained through resident cost minimization modeling, and a resident response model is obtained according to the resident cost objective function and resident response decision variables; Obtaining market electricity price information, obtaining a load aggregator profit objective function through load aggregator profit maximization modeling based on the market electricity price information, and obtaining a load aggregator regulation model based on the load aggregator profit objective function and load aggregator regulation decision variables; A resident electricity price incentive response model is obtained by combining the resident response model and the load aggregator regulation model. The load aggregator's day-ahead electricity price is obtained by solving the resident electricity price incentive response model through dual-objective optimization. Load aggregator regulation is achieved through electricity price pricing based on the load aggregator's day-ahead electricity price.

2. The community-level load aggregator control method according to claim 1, characterized in that: The obtaining of community residents' electricity consumption data and obtaining residents' electricity load information through a residents' electricity load identification model according to the community residents' electricity consumption data include: Collecting community residents' electricity consumption data through smart meters and eliminating dimensions through data normalization, wherein the community residents' electricity consumption data includes active power of residents' electricity consumption and effective value of residents' electricity consumption current; According to the community residents' electricity consumption data, the load equipment operation objective function is solved by an improved GWO algorithm to obtain the residents' electricity load information.

3. The community-level load aggregator control method according to claim 2, characterized in that: The obtaining of residential electricity load information by solving the load equipment operation objective function through the improved GWO algorithm according to the community residents' electricity consumption data includes: Initializing a population, wherein the load equipment operation objective function values ​​of individuals in the population are used as fitness values; Update the individual positions of the population through differential perturbations; The GWO algorithm is continuously iterated to select the individual with the highest fitness value as the optimal solution, that is, the residents' electricity load information.

4. The community-level load aggregator control method according to claim 1, characterized in that: The resident cost objective function obtained by resident cost minimization modeling according to the resident electricity load information includes: Preset the time interval for power load transfer; Preset transferable load power range; The mathematical expression of the resident cost objective function is: C=α(t4-t2)Q Among them, M t is the real-time electricity price per unit time the day before, u is the basic load of residential electricity consumption, is the activation variable of device j of resident i at time t, is the j-device power of resident i, the original operation time period of the transferable load is [t1, t2], the operation time period after the transferable load is transferred is [t3, t4], α is the resident comfort cost per unit time, Q is the total transfer power of the transferable load, C is the resident comfort cost, and Fa is the resident cost objective function; The constraints of the resident cost objective function are that the transferred load must be within the electricity load transfer time interval, the power after the load transfer must be within the transferable load power interval, and the required electric energy before and after the load transfer remains unchanged. The optimized load curve value is not less than 0, and the minimum load after residents participate in demand response is not lower than the minimum load before participating in demand response.

5. The community-level load aggregator control method according to claim 1, characterized in that: The load aggregator profit objective function obtained by load aggregator profit maximization modeling according to the market electricity price information includes: Preset the real-time electricity price range for the day before; The mathematical expression of the load aggregator's revenue objective function is: Fb=max{H-cost1-cost2-Cr} Among them, cost1 is the electricity purchase cost of the load aggregator, cost2 is the equipment loss cost of the load aggregator, Cr is the risk cost of the load aggregator, Fb is the load aggregator's profit objective function, H is the load aggregator's power output profit, M t is the real-time electricity price per unit at the day before, and B is the resident response decision variable; The constraints of the load aggregator's revenue objective function are that the real-time electricity price per unit time the day before must be within the real-time electricity price range per unit time the day before, the electricity price the day before is equal to the average electricity price the day before, and the load power remains balanced.

6. The community-level load aggregator control method according to claim 1, characterized in that: The day-ahead electricity price of the load aggregator obtained by solving the residential electricity price incentive response model through dual-objective optimization includes: Obtaining resident response decision variable values ​​through a Cplex solver according to the resident response model; The load aggregator control model is solved by an improved genetic algorithm according to the resident response decision variable value to obtain the load aggregator's day-ahead electricity price.

7. The community-level load aggregator control method according to claim 6, characterized in that: The method of solving the load aggregator regulation model by using an improved genetic algorithm according to the resident response decision variable value to obtain the load aggregator's day-ahead electricity price includes: The resident response decision variable value is substituted into the load aggregator profit objective function as a known condition; Initialize the population through Logistic chaos mapping; The load aggregator profit objective function value of the individuals in the population is taken as the fitness value, and the individual with the highest fitness value is selected through iterative selection of the genetic algorithm as the optimal solution, that is, the day-ahead electricity price of the load aggregator.

8. A community-level load aggregator control system based on machine learning, characterized in that: The system is applied to the community-level load aggregator control method as described in any one of claims 1 to 7, including a data acquisition module, a residential electricity load identification module, a residential cost modeling module, a load aggregator revenue modeling module, a residential electricity price incentive response module, and an electricity price pricing and control module; The data acquisition module is used to obtain community residents' electricity consumption data and market electricity price information; The residential power load identification module is used to obtain residential power load information based on community residents' power consumption data through a residential power load identification model; The resident cost modeling module is used to construct a resident response model based on the community residents' electricity consumption data through the resident cost objective function and the resident response decision variables; The load aggregator revenue modeling module is used to construct a load aggregator regulation model based on market electricity price information through a load aggregator revenue objective function and a load aggregator regulation decision variable; The residential electricity price incentive response module is used to combine the resident response model and the load aggregator regulation model to form a residential electricity price incentive response model, and determine the load aggregator's day-ahead electricity price through a dual-objective optimization solution to balance the minimization of resident costs and the maximization of load aggregator benefits; The electricity price pricing and regulation module is used to set electricity prices according to the optimized day-ahead electricity price of the load aggregator and implement load aggregator regulation through a feedback mechanism to encourage residents to use electricity during off-peak hours and optimize the load curve.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, it implements the community-level load aggregator control method as described in any one of claims 1-7.

10. A storage medium containing computer executable instructions, characterized in that: The computer executable instructions, when executed by a computer processor, are used to execute the community-level load aggregator regulation method as described in any one of claims 1-7.

Images