Multi-agent cooperative loss reduction multi-objective optimization method and system based on harmony algorithm

Abstract

The invention provides a harmony algorithm-based multi-agent cooperative loss reduction multi-objective optimization method and system. The method comprises the following steps of: obtaining operation data of a power distribution network; based on the operation data, a harmony algorithm is utilized to solve a pre-constructed multi-subject optimization loss reduction multi-target model, and a power distribution network operation scheme is obtained; performing cooperative loss reduction on each income main body in the power distribution network by using the operation scheme of the power distribution network; wherein the multi-subject optimization loss reduction multi-objective model is constructed by taking maximization of the comprehensive net income of each income subject, the purchase energy satisfaction degree of the user and the new energy consumption rate as objectives; according to the method, different profit subjects are considered when the model is constructed, so that the profit appeals of different subjects are reflected, and the influence of multi-subject cooperative operation on loss reduction optimization of the power distribution network in the market is also reflected; and the model is solved by adopting a harmony algorithm, so that the optimization capacity of the multi-target type model is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of multi-objective optimization of coordinated loss reduction in distribution networks, and in particular to a multi-agent coordinated loss reduction multi-objective optimization method and system based on a harmony algorithm. Background Art

[0002] At present, in the market environment, the issue of distribution network line loss has attracted more and more attention from various entities. Different entities, such as distributed power generation companies, load users, integrated energy service providers, and electric vehicle charging stations, should bear corresponding responsibilities for distribution network line loss. The behaviors and decisions of these entities directly affect the operation and line loss of the distribution network. Therefore, it is crucial to conduct in-depth research on the role of different entities in reducing distribution network losses and optimizing operation.

[0003] In the current electricity market, all stakeholders start from their own interests. In order to maximize their own interests, they usually do not consider the overall line loss situation. How to coordinate operation to play a beneficial role in reducing network losses has been less studied. For the loss reduction optimization operation of traditional distribution networks, it mainly stays at the loss reduction optimization for a single entity, or considers the loss reduction optimization in a single group, without considering the impact of multiple entities in the market and the coordination between multiple entities. Summary of the invention

[0004] In order to solve the problem that the prior art only performs loss reduction optimization on a single subject without considering the influence of multiple subjects in the market, and that the conventional optimization algorithm used in the prior art is prone to fall into local optimum and slow convergence when solving multi-objective models, the present invention proposes a multi-agent collaborative loss reduction multi-objective optimization method based on a harmony algorithm, comprising:

[0005] Obtaining the operation data of the distribution network;

[0006] Based on the operation data, a pre-built multi-agent optimization loss reduction multi-objective model is solved using a harmony algorithm to obtain a distribution network operation plan;

[0007] Using the distribution network operation plan, collaboratively reducing losses for various revenue entities in the distribution network;

[0008] Among them, the multi-agent optimization loss reduction multi-objective model is constructed with the goals of maximizing the comprehensive net income of each benefit entity, the user's energy purchase satisfaction and the new energy consumption rate. The benefit entities include at least several of the following: distribution companies, integrated energy service providers, and distributed power generation companies.

[0009] Optionally, the multi-agent optimization loss reduction multi-objective model includes the following construction process:

[0010] Calculate the comprehensive net income of each income subject based on the income from energy sales of each income subject and the total operating costs corresponding to each income subject;

[0011] Calculating the user's energy purchase satisfaction according to the user's purchased energy amount and the preference coefficient corresponding to the purchased energy amount;

[0012] Calculate the new energy consumption rate based on the new energy consumption and actual power generation;

[0013] The objective function is constructed with the goal of maximizing the comprehensive net income of each income subject, the user's energy purchase satisfaction and the new energy consumption rate;

[0014] According to the objective function, formulate constraint conditions corresponding to the objective function;

[0015] Based on the objective function and the constraints, a multi-agent optimization and loss reduction multi-objective model is constructed.

[0016] Optionally, the total operating cost of the distribution company includes one or more of the following: energy purchase cost of the distribution company, operation and maintenance cost, cost of electricity purchased by the distribution company from the superior main grid, cost of electricity purchased by the distribution company from distributed power generation companies, environmental cost of the distribution company and network loss cost;

[0017] The total operating cost of the integrated energy service provider includes one or more of the following: energy purchase cost of the integrated energy service provider, investment cost, environmental cost of the integrated energy service provider, operation and maintenance cost, and network loss cost;

[0018] The income from energy sales of the distributed power generation generator includes one or more of the following: income from energy sales of the distribution company and income from photovoltaic grid-connected subsidies;

[0019] The total operating cost of the distributed power generator includes one or more of the following: unit operation and maintenance cost and unit power generation fuel cost;

[0020] The amount of energy purchased includes one or more of the following: the amount of heat purchased and the amount of electricity purchased;

[0021] The actual power generation includes one or more of the following: actual photovoltaic power generation and actual wind power generation;

[0022] The constraint conditions include one or more of the following: power balance constraint, branch voltage constraint, upper power grid power purchase constraint, real-time electricity price constraint, power constraint, power purchase and sale constraint and distribution network power flow equation constraint.

[0023] Optionally, the expression corresponding to the objective function is as follows:

[0024]

[0025] Among them, B DC represents the comprehensive net income of the distribution company; F IES F represents the comprehensive net income of the integrated energy service provider; DG F represents the comprehensive net income of distributed power generators; user Indicates the user's purchasing satisfaction; F RES represents the new energy consumption rate; B sell Represents the revenue from selling electricity to distribution companies; C In represents the energy purchase cost of the distribution company; C OM represents the operation and maintenance cost; C Buy.grid C represents the cost of electricity purchased by the distribution company from the upper grid; buy.DG E represents the cost of electricity purchased by the distribution company from the distributed generation generator; DC represents the environmental cost of the distribution company; C Loss represents the network loss cost; C Sell Represents the income from energy sales of integrated energy service providers; C buy represents the energy purchase cost of the integrated energy service provider; C Cap represents the investment cost; represents the environmental cost of the integrated energy service provider; R Sell Represents the revenue of the distribution company from energy sales; R Gusb Represents the income from photovoltaic grid-connected subsidies; C Oper represents the unit operation and maintenance cost of the distributed generation generator; C Cgen represents the unit power generation fuel cost of distributed generation generators; v e Indicates the first preference coefficient corresponding to the amount of electricity purchased by the user; represents the amount of electricity purchased by the user during period t; αe represents the second preference coefficient corresponding to the amount of electricity purchased by the user; a′ represents the constant coefficient; v h Indicates the first preference coefficient corresponding to the user's heat purchase; α h Indicates the second preference coefficient corresponding to the user's purchase of heat; Indicates the amount of heat purchased by users during period t; represents the actual photovoltaic power generation in the hth hour; represents the actual wind power generation in hour h; represents the photovoltaic power generation purchased by the integrated energy service provider from the distributed power generation company at hour h; It represents the amount of wind power purchased by the integrated energy service provider from the distributed power generation company at the hth hour.

[0026] Optionally, based on the operation data, a pre-built multi-agent optimization loss reduction multi-objective model is solved using a harmony algorithm to obtain a distribution network operation plan, including:

[0027] Initializing a harmony memory library and harmony parameters, wherein the harmony parameters include: a harmony memory library, a harmony memory consideration rate, and a number of iterations, wherein each initial harmony in the harmony memory library is the operating data of the distribution network, and the fitness function of the harmony is represented by the objective function;

[0028] The initial harmony memory consideration rate is updated by using a dynamic adjustment calculation formula to obtain an updated harmony memory consideration rate; a new harmony is created based on each initial harmony according to the updated harmony memory consideration rate, and a fitness function value corresponding to the new harmony is calculated according to the fitness function;

[0029] Based on the fitness function value corresponding to the new harmony and the fitness function value corresponding to the initial harmony, the harmony memory library is iteratively updated using an elite retention strategy until the number of iterations is reached; the harmony memory library updated when the number of iterations is reached is used as the final harmony memory library;

[0030] Using intuitive fuzzy theory and the approximate ideal solution sorting method, the closeness evaluation value of each harmony in the final harmony memory library is calculated, and the harmony with the highest closeness evaluation value is selected as the distribution network operation plan;

[0031] The operation data of the distribution network includes one or more of the following: output power of distributed generators, load distribution ratio and energy trading strategy of energy service providers.

[0032] Optionally, the dynamic adjustment calculation formula is as follows:

[0033]

[0034] Where HMCR(t0) represents the harmonic memory consideration rate at the t0th iteration; HMCR Min Indicates the preset minimum threshold of harmony memory consideration rate; HMCR Max Indicates the preset maximum threshold of harmony memory consideration rate; T Max Indicates the maximum number of iterations.

[0035] Optionally, the step of creating a new harmony based on each initial harmony according to the updated harmony memory consideration rate comprises:

[0036] If the updated harmony memory consideration rate is greater than a randomly generated random number, randomly selecting a harmony from the harmony memory library as a newly created harmony;

[0037] If the updated harmony memory consideration rate is not greater than the random number, a harmony is randomly generated from a preset solution space as a created new harmony.

[0038] Optionally, based on the fitness function value corresponding to the new harmony and the fitness function value corresponding to the initial harmony, the harmony memory library is updated using an elite retention strategy, including:

[0039] The harmony with the smallest fitness function value among the initial harmony is used as the underlying harmony;

[0040] The fitness function value of the underlying harmony is compared with the fitness function value corresponding to the new harmony. If the fitness function value of the underlying harmony is less than the fitness function value corresponding to the new harmony, the new harmony is used as a new underlying harmony, and the harmony memory bank is updated according to the new underlying harmony.

[0041] Optionally, the use of the intuitive fuzzy theory and the approximate ideal solution sorting method to calculate the closeness evaluation value of each harmony in the final harmony memory library includes:

[0042] According to the final harmony memory library, a decision matrix is ​​established by using an intuitive fuzzy theory method, and the membership, non-membership and hesitation of the decision matrix are calculated;

[0043] Based on the membership, non-membership and hesitation of the decision matrix, the closeness evaluation value of each harmony in the final harmony memory library is calculated using the approximate ideal solution sorting method.

[0044] Optionally, the method of calculating the closeness evaluation value of each harmony in the final harmony memory library based on the membership, non-membership and hesitation of the decision matrix and using the approximate ideal solution sorting method includes:

[0045] According to the membership, non-membership and hesitation of the decision matrix, the distance between each harmony in the final harmony memory bank and the preset theoretical optimal solution and theoretical worst solution is calculated by using the approximate ideal solution sorting method;

[0046] According to the distances between each harmony and the theoretical optimal solution and the theoretical worst solution, respectively, a closeness evaluation value corresponding to each harmony in the final harmony memory library is calculated.

[0047] Based on the same inventive concept, the present invention also provides a multi-agent collaborative loss reduction multi-objective optimization system based on a harmony algorithm, comprising:

[0048] A data acquisition module, used to acquire the operation data of the distribution network;

[0049] A model solving module, for solving a pre-built multi-agent optimization loss reduction multi-objective model based on the operation data using a harmony algorithm to obtain a distribution network operation plan;

[0050] A collaborative loss reduction module, used to collaboratively reduce losses for each income subject in the distribution network according to the distribution network operation plan;

[0051] Among them, the multi-agent optimization loss reduction multi-objective model is constructed with the goals of maximizing the comprehensive net income of each benefit entity, the user's energy purchase satisfaction and the new energy consumption rate. The benefit entities include at least multiple of the following: distribution companies, integrated energy service providers, and distributed power generation companies.

[0052] Optionally, the optimization system further includes: a model building module, which is used to:

[0053] Calculate the comprehensive net income of each income subject based on the income from energy sales of each income subject and the total operating costs corresponding to each income subject;

[0054] Calculating the user's energy purchase satisfaction according to the user's purchased energy amount and the preference coefficient corresponding to the purchased energy amount;

[0055] Calculate the new energy consumption rate based on the new energy consumption and actual power generation;

[0056] The objective function is constructed with the goal of maximizing the comprehensive net income of each income subject, the user's energy purchase satisfaction and the new energy consumption rate;

[0057] According to the objective function, formulate constraint conditions corresponding to the objective function;

[0058] Based on the objective function and the constraints, a multi-agent optimization and loss reduction multi-objective model is constructed.

[0059] Optionally, the total operating cost of the distribution company includes one or more of the following: energy purchase cost of the distribution company, operation and maintenance cost, cost of electricity purchased by the distribution company from the superior main grid, cost of electricity purchased by the distribution company from distributed power generation companies, environmental cost of the distribution company and network loss cost;

[0060] The total operating cost of the integrated energy service provider includes one or more of the following: energy purchase cost of the integrated energy service provider, investment cost, environmental cost of the integrated energy service provider, operation and maintenance cost, and network loss cost;

[0061] The income from energy sales of the distributed power generation generator includes one or more of the following: income from energy sales of the distribution company and income from photovoltaic grid-connected subsidies;

[0062] The total operating cost of the distributed power generator includes one or more of the following: unit operation and maintenance cost and unit power generation fuel cost;

[0063] The amount of energy purchased includes one or more of the following: the amount of heat purchased and the amount of electricity purchased;

[0064] The actual power generation includes one or more of the following: actual photovoltaic power generation and actual wind power generation;

[0065] The constraint conditions include one or more of the following: power balance constraint, branch voltage constraint, upper power grid power purchase constraint, real-time electricity price constraint, power constraint, power purchase and sale constraint and distribution network power flow equation constraint.

[0066] Optionally, the expression corresponding to the objective function is as follows:

[0067]

[0068] Among them, B DC represents the comprehensive net income of the distribution company; F IES F represents the comprehensive net income of the integrated energy service provider; DG F represents the comprehensive net income of distributed power generators; user Indicates the user's purchasing satisfaction; F RES represents the new energy consumption rate; B sell Represents the revenue from selling electricity to distribution companies; C In represents the energy purchase cost of the distribution company; C OM represents the operation and maintenance cost; C Buy.grid C represents the cost of electricity purchased by the distribution company from the upper grid; buy.DG E represents the cost of electricity purchased by the distribution company from the distributed generation generator; DC represents the environmental cost of the distribution company; C Loss represents the network loss cost; C Sell Represents the income from energy sales of integrated energy service providers; C buy represents the energy purchase cost of the integrated energy service provider; C Cap represents the investment cost; represents the environmental cost of the integrated energy service provider; R Sell Represents the revenue of the distribution company from energy sales; R Gusb Represents the income from photovoltaic grid-connected subsidies; C Oper represents the unit operation and maintenance cost of the distributed generation generator; C Cgen represents the unit power generation fuel cost of distributed generation generators; v e Indicates the first preference coefficient corresponding to the amount of electricity purchased by the user; represents the amount of electricity purchased by users during period t; α e represents the second preference coefficient corresponding to the amount of electricity purchased by the user; a′ represents the constant coefficient; v h Indicates the first preference coefficient corresponding to the user's heat purchase; α h Indicates the second preference coefficient corresponding to the user's purchase of heat; Indicates the amount of heat purchased by users during period t; represents the actual photovoltaic power generation in the hth hour; represents the actual wind power generation in hour h; represents the photovoltaic power generation purchased by the integrated energy service provider from the distributed power generation company at hour h; It represents the amount of wind power purchased by the integrated energy service provider from the distributed power generation company at the hth hour.

[0069] Optionally, the model solving module includes:

[0070] A parameter initialization submodule is used to initialize the harmony memory library and harmony parameters, wherein the harmony parameters include: harmony memory library, harmony memory consideration rate, and number of iterations, wherein each initial harmony in the harmony memory library is the operating data of the distribution network, and the fitness function of the harmony is represented by the objective function;

[0071] A harmony creation submodule is used to update the initial harmony memory consideration rate using a dynamic adjustment calculation formula to obtain an updated harmony memory consideration rate; create a new harmony based on each initial harmony according to the updated harmony memory consideration rate, and calculate the fitness function value corresponding to the new harmony according to the fitness function;

[0072] A memory library updating submodule, for iteratively updating the harmony memory library based on the fitness function value corresponding to the new harmony and the fitness function value corresponding to the initial harmony, using an elite retention strategy, until the number of iterations is reached; and using the harmony memory library updated when the number of iterations is reached as the final harmony memory library;

[0073] A closeness evaluation submodule is used to calculate the closeness evaluation value of each harmony in the final harmony memory library by using intuitive fuzzy theory and approximate ideal solution sorting method, and select the harmony with the highest closeness evaluation value as the distribution network operation plan;

[0074] The operation data of the distribution network includes one or more of the following: output power of distributed generators, load distribution ratio and energy trading strategy of energy service providers.

[0075] Optionally, the dynamic adjustment calculation formula is as follows:

[0076]

[0077] Where HMCR(t0) represents the harmonic memory consideration rate at the t0th iteration; HMCR Min Indicates the preset minimum threshold of harmony memory consideration rate; HMCR Max Indicates the preset maximum threshold of harmony memory consideration rate; T Max Indicates the maximum number of iterations.

[0078] Optionally, the harmony creation submodule includes:

[0079] an initial selection unit, configured to randomly select a harmony from the harmony memory library as a newly created harmony when the updated harmony memory consideration rate is greater than a randomly generated random number;

[0080] The space generation unit is used to randomly generate a harmony from a preset solution space as a created new harmony when the updated harmony memory consideration rate is not greater than the random number.

[0081] Optionally, the memory library update submodule includes:

[0082] A bottom layer setting unit, used for taking the harmony with the smallest fitness function value among the initial harmony as the bottom layer harmony;

[0083] The bottom-level replacement unit is used to compare the fitness function value of the bottom-level harmony with the fitness function value corresponding to the new harmony. If the fitness function value of the bottom-level harmony is less than the fitness function value corresponding to the new harmony, the new harmony is used as a new bottom-level harmony, and the harmony memory bank is updated according to the new bottom-level harmony.

[0084] Optionally, the closeness evaluation submodule includes:

[0085] A matrix generating unit, used for establishing a decision matrix according to the final harmony memory library by using an intuitive fuzzy theory method, and calculating the membership, non-membership and hesitation of the decision matrix;

[0086] The closeness calculation unit is used to calculate the closeness evaluation value of each harmony in the final harmony memory library based on the membership, non-membership and hesitation of the decision matrix and using the approximate ideal solution sorting method.

[0087] Optionally, the proximity calculation unit includes:

[0088] A distance calculation subunit, for calculating the distance between each harmony in the final harmony memory bank and the preset theoretical optimal solution and theoretical worst solution respectively, according to the membership, non-membership and hesitation of the decision matrix, using an approximate ideal solution sorting method;

[0089] The evaluation unit is used to calculate the closeness evaluation value corresponding to each harmony in the final harmony memory library according to the distance between each harmony and the theoretical optimal solution and the theoretical worst solution respectively.

[0090] In another aspect, the present invention further provides an electronic device, comprising: at least one processor and a memory; the memory and the processor are connected via a bus;

[0091] The memory is used to store one or more programs;

[0092] When the one or more programs are executed by the at least one processor, a multi-agent collaborative loss reduction multi-objective optimization method based on a harmony algorithm as described above is implemented.

[0093] On the other hand, the present invention also provides a computer-readable storage medium having an execution program stored thereon, which, when executed, implements the multi-agent collaborative loss reduction multi-objective optimization method based on the harmony algorithm as described above.

[0094] Compared with the prior art, the present invention has the following beneficial effects:

[0095] The present invention provides a multi-agent collaborative loss reduction multi-objective optimization method and system based on a harmony algorithm, comprising: obtaining operation data of a distribution network; based on the operation data, solving a pre-constructed multi-agent optimization loss reduction multi-objective model using a harmony algorithm to obtain an operation plan for the distribution network; using the distribution network operation plan to collaboratively reduce losses for each benefit subject in the distribution network; wherein the multi-agent optimization loss reduction multi-objective model is constructed with the goal of maximizing the comprehensive net income of each benefit subject, the user's energy purchase satisfaction and the new energy consumption rate, and the benefit subjects include at least a plurality of the following: a distribution company, an integrated energy service provider, and a distributed power generation company; when constructing the model, the present application considers different benefit subjects, which not only reflects the interests of different subjects, but also reflects the impact of the collaborative operation of multiple subjects in the market on the loss reduction optimization of the distribution network; and by solving the model using a harmony algorithm, it is beneficial to improve the optimization ability of multi-objective type models. BRIEF DESCRIPTION OF THE DRAWINGS

[0096] Figure 1 A schematic diagram of a flow chart of a multi-agent collaborative loss reduction multi-objective optimization method based on a harmony algorithm provided by the present invention;

[0097] Figure 2 A schematic diagram of the execution flow of the harmony algorithm in a multi-agent collaborative loss reduction multi-objective optimization method based on the harmony algorithm provided by the present invention;

[0098] Figure 3 A schematic diagram of the structure of a multi-agent collaborative loss reduction multi-objective optimization system based on a harmony algorithm provided by the present invention;

[0099] Figure 4 The present invention provides a schematic structural diagram of an electronic device. DETAILED DESCRIPTION

[0100] The present invention proposes a multi-agent collaborative loss reduction multi-objective optimization method, system, device and medium based on a harmony algorithm. The specific implementation methods of the present invention are further described in detail below in conjunction with the accompanying drawings.

[0101] Embodiment 1:

[0102] The present invention provides a multi-agent collaborative loss reduction multi-objective optimization method based on a harmony algorithm, and the flow chart is as follows: Figure 1 As shown, including:

[0103] Step 1: Obtain the operation data of the distribution network;

[0104] Step 2: Based on the operation data, the pre-built multi-agent optimization loss reduction multi-objective model is solved using the harmony algorithm to obtain the distribution network operation plan;

[0105] Step 3: Use the distribution network operation plan to coordinate loss reduction for each income entity in the distribution network;

[0106] Among them, the multi-agent optimization loss reduction multi-objective model is constructed with the goals of maximizing the comprehensive net income of each benefit entity, the user's energy purchase satisfaction and the new energy consumption rate. Each benefit entity includes at least several of the following: distribution companies, integrated energy service providers, and distributed power generation companies.

[0107] Generally, in a competitive electricity market, there are several main entities in the entire distribution system: distribution companies, integrated energy service providers, distributed power generators, prosumers and load users. Distributed power generators sell the generated electricity to distribution companies and integrated energy service providers at a certain price; distribution companies sell electricity to integrated energy service providers and users based on real-time electricity prices; integrated energy service providers sell electricity, heat and cold energy to users; load users purchase electricity from different income entities. Since the main goal of prosumers is not to make a profit, the net income of users and prosumers is not considered here. Different income entities all proceed from their own interests and maximize their own interests. The overall line loss is usually not considered, which can easily lead to low grid efficiency and reduced power quality. To solve this problem, the present invention first obtains the operating data of the distribution network, and based on the operating data of the distribution network, considers the economic, technical, environmental protection and other aspects of different benefit entities to optimize the loss reduction multi-objective model, thereby reflecting the interests of different benefit entities and promoting loss reduction and efficiency improvement. For example, the operating data of the distribution network can include: the output power of distributed generators, the load distribution ratio and the energy trading strategy of energy service providers, wherein the output power and load distribution ratio of distributed generators can be obtained through data acquisition and monitoring control (Supervisory The SCADA system collects data, which can collect various operating parameters in the distribution network, such as voltage, current, power, etc. in real time, and transmit these data to the central control room for analysis and processing. For information such as the output power of distributed generators and the load distribution ratio, the SCADA system can provide effective data support; the energy trading strategy of integrated energy service providers involves business decisions and market operations, so the method of obtaining such information is usually different from public technical data or physical measurements. It can be obtained through annual reports, quarterly reports or other types of announcements published by large energy service providers on their official websites. The present invention optimizes the operating data of the above-mentioned distribution network, determines the distribution network operation plan that takes into account multiple benefit entities, and realizes the coordinated loss reduction of the distribution network. Specifically:

[0108] In one implementation, the above-mentioned multi-agent optimization loss reduction multi-objective model may include the following construction process:

[0109] Calculate the comprehensive net income of each income subject based on the income from energy sales and the total operating costs of each income subject;

[0110] Calculate the user's energy purchase satisfaction based on the user's purchased energy amount and the preference coefficient corresponding to the purchased energy amount;

[0111] Calculate the new energy consumption rate based on the new energy consumption and actual power generation;

[0112] The objective function is constructed with the goal of maximizing the comprehensive net income of each income subject, the user's energy purchase satisfaction and the new energy consumption rate;

[0113] According to the objective function, formulate the constraint conditions corresponding to the objective function;

[0114] Based on the objective function and constraints, a multi-agent optimization loss reduction multi-objective model is constructed;

[0115] In this implementation, by accurately calculating the income from energy sales and the total operating cost of each income subject, the comprehensive net income is obtained, and this is used as the optimization goal, which can make the energy allocation and pricing strategies more reasonable, which is conducive to motivating the income subject to improve the efficiency of energy production and utilization, reduce operating costs, and thus improve the economic benefits of each income subject; the energy purchase satisfaction is calculated according to the user's purchased energy volume and preference coefficient, which can reflect the user's satisfaction and loyalty to the energy supply. By optimizing the energy supply strategy, the user's personalized needs can be met, and user satisfaction can be improved, which is conducive to enhancing user stickiness and expanding market share; by calculating the ratio of new energy consumption and actual power generation, the utilization efficiency of new energy can be evaluated, and the goal of maximizing the new energy consumption rate can be promoted. The energy system can be more flexibly dispatched and allocated for renewable energy, wind and solar power abandonment can be reduced, and the utilization rate and economic benefits of renewable energy can be improved. This implementation method comprehensively considers the economic benefits of various beneficiaries, users' satisfaction with energy purchases, and the new energy consumption rate, and considers the diversification of subjects involved in collaborative optimization of loss reduction. A multi-objective model that reflects the interests of different subjects is selected to achieve the goal of comprehensive optimization between maximizing the interests of different subjects and reducing network losses. A multi-subject optimization loss reduction multi-objective model that can reflect the mutual promotion and constraint between different subjects and loss reduction is constructed. Energy distribution, pricing strategies, and dispatching methods can be optimized to reduce energy losses, improve the overall operating efficiency and reliability of the energy system, and thus achieve global optimization of the energy system.

[0116] In this implementation, the comprehensive net income of the distribution company (for example, B DC can be expressed as the revenue from selling electricity to the distribution company (for example, B sell The difference between the total cost of operation and the total cost of operation; For example, the total cost of operation of the distribution company may include one or more of the following: the cost of purchasing energy for the distribution company (for example, C In ), operation and maintenance costs (for example, C OM The cost of electricity purchased by the distribution company from the upper grid (for example, C Buy.grid The cost of electricity purchased by the distribution company from the distributed power generation company (for example, C buy.DG ), distribution company environmental costs (for example, EDC ) and network loss costs (e.g., C Loss express);

[0117] In this implementation, the comprehensive net income of the integrated energy service provider (for example, F IES can be expressed as the revenue from energy sales of the integrated energy service provider (for example, C Sell ) and the difference between the total operating cost; For example, the total operating cost of the integrated energy service provider may include one or more of the following: the energy purchase cost of the integrated energy service provider (for example, C buy Indicated), investment cost (for example, C Cap ), comprehensive energy service provider environmental cost (for example, C OM ), operation and maintenance costs (for example, represents) and network loss cost C Loss ;

[0118] In this implementation, the comprehensive net income of the distributed power generator (for example, F DG The revenue from energy sales of distributed power generators can be expressed as the difference between the revenue from energy sales of distributed power generators and the total operating cost. For example, the revenue from energy sales of distributed power generators can include one or more of the following: the revenue from energy sales of distribution companies (for example, R Sell ) and PV grid-connected subsidy income (for example, R Gusb The total operating cost of a distributed power generator may include one or more of the following: unit operation and maintenance cost (for example, C Oper ) and unit fuel cost for electricity generation (e.g., C Cgen express);

[0119] In this implementation, the user's energy purchase satisfaction (for example, F user It represents the satisfaction degree of the user when purchasing electric energy and thermal energy, that is, the user's utility function, which is usually non-decreasing and convex, and has several forms such as quadratic form and logarithmic form. In this implementation, the quadratic form is used to represent it, and its expression corresponds to the product of the user's purchased energy amount and the preference coefficient corresponding to the purchased energy amount; for example, the purchased energy amount may include one or more of the following: purchased heat amount and purchased electricity amount;

[0120] In this implementation, the new energy consumption rate can be expressed as the ratio between the consumption of new energy and the actual power generation; for example, the consumption of new energy can include one or more of the following: photovoltaic power generation and wind power generation purchased by the integrated energy service provider from the distributed power generation company; the actual power generation can include one or more of the following: actual photovoltaic power generation and actual wind power generation; when constructing the objective function according to the consumption of new energy in the integrated energy system, the consumption rate of new energy can be obtained by calculating the ratio between the consumption of new energy by the integrated energy service provider and the actual new energy power generation by the distributed power generation company;

[0121] For example, the expression corresponding to the above objective function can be as follows:

[0122]

[0123] Among them, B DC represents the comprehensive net income of the distribution company; F IES F represents the comprehensive net income of the integrated energy service provider; DG F represents the comprehensive net income of distributed power generators; user Indicates the user's purchasing satisfaction; F RES represents the new energy consumption rate; B sell Represents the revenue from selling electricity to distribution companies; C In represents the energy purchase cost of the distribution company; C OM represents the operation and maintenance cost; C Buy.grid C represents the cost of electricity purchased by the distribution company from the upper grid; buy.DG E represents the cost of electricity purchased by the distribution company from the distributed generation generator; DC represents the environmental cost of the distribution company; C Loss represents the network loss cost; C Sell Represents the income from energy sales of integrated energy service providers; C buy represents the energy purchase cost of the integrated energy service provider; C Cap represents the investment cost; represents the environmental cost of the integrated energy service provider; R Sell Represents the revenue of the distribution company from energy sales; R Gusb Represents the income from photovoltaic grid-connected subsidies; C Oper represents the unit operation and maintenance cost of the distributed generation generator; C Cgen represents the unit power generation fuel cost of distributed generation generators; v e Indicates the first preference coefficient corresponding to the amount of electricity purchased by the user; represents the amount of electricity purchased by users during period t; α e represents the second preference coefficient corresponding to the amount of electricity purchased by the user; a′ represents the constant coefficient; v h Indicates the first preference coefficient corresponding to the user's heat purchase; αh Indicates the second preference coefficient corresponding to the user's purchase of heat; Indicates the amount of heat purchased by users during period t; represents the actual photovoltaic power generation in the hth hour; represents the actual wind power generation in hour h; represents the photovoltaic power generation purchased by the integrated energy service provider from the distributed power generation company at hour h; represents the wind power purchased by the integrated energy service provider from the distributed power generator in the hth hour; in the energy transaction process, the power sales company or the integrated energy service provider formulates a price strategy based on the load demand, and the user side responds to the demand based on the price information. In this example, the objective function of the integrated energy service provider is established with the goal of maximizing the comprehensive net profit of the integrated energy service provider. Specifically, the comprehensive net profit F of the integrated energy service provider is IES It can be expressed as the income C obtained by the integrated energy service provider from selling energy Sell and the total operating cost corresponding to the integrated energy service provider, where the total operating cost corresponding to the integrated energy service provider may at least include: the energy purchase cost C of the integrated energy service provider buy 、Investment cost C Cap , Environmental costs of integrated energy service providers Operation and maintenance cost C OM and network loss cost C Loss Etc.; In addition, the objective function can be understood in two layers, where the first layer is to maximize the comprehensive net profit of the distribution company, and the second layer is to respectively target the comprehensive net profit of the integrated energy service provider, the comprehensive net profit of the distributed power generation company, the user's energy purchase satisfaction, and the maximum new energy consumption rate. By dividing the objective function into two layers, the optimization objectives of the constructed multi-agent optimization loss reduction multi-objective model can be made clearer and more structured. In the two-layer objective function, the first-layer objective (i.e., the maximum comprehensive net profit of the distribution company) is regarded as a higher priority objective, while the second-layer objective (i.e., the comprehensive net profit of the integrated energy service provider and the distributed power generation company, the user's energy purchase satisfaction, and the maximum new energy consumption rate) is regarded as a secondary objective. This hierarchical structure is conducive to clarifying priorities in the decision-making process, ensuring that while pursuing overall optimization, the interests of all parties can be met, and the hierarchical objective function is conducive to promoting coordination among different benefit subjects. For example, in the process of solving the model, the first-layer objective can be met first, and then the second-layer objective can be optimized as much as possible on the premise of meeting this objective. This coordination mechanism is conducive to achieving a win-win situation among multiple benefit subjects. In some cases, the first-layer target may also be referred to as the upper-layer target in a double-layer target, and the second-layer target may also be referred to as the lower-layer target in a double-layer target.

[0124] In this example, the above distribution company's comprehensive net income BDC The network loss cost C is calculated as Loss The corresponding expression can be as follows:

[0125]

[0126] Among them, M d represents the number of the dth typical day in a year; d = 1...D; D represents the total number of typical days; C Loss.d.t represents the network loss cost of the integrated energy service provider at time t on the dth typical day; c Loss represents the unit network loss cost; P Loss.d.t represents the network loss cost of all branches of the integrated energy service provider at time t on the dth typical day; R l.d.t represents the resistance of branch l of the integrated energy service provider at time t on the dth typical day; l=1...L; L represents the total number of branches; P l.d.t represents the active power flowing into the terminal node from branch l at time t on the dth typical day of the integrated energy service provider; Q l.d.t represents the reactive power of branch l flowing into the terminal node at time t on the dth typical day of the integrated energy service provider; U l.d.t represents the node voltage of the terminal node of branch l of the integrated energy service provider at time t on the dth typical day; Δt represents the time interval; in this example, by refining the network loss cost to each moment of each typical day and each branch, the cost incurred by the distribution company due to network loss during operation can be more accurately measured, and the time interval Δt is considered in the expression, so that the network loss cost can dynamically reflect the loss of the power network in different time periods, which is conducive to the distribution company to more accurately grasp the changing trend of network loss and formulate energy allocation and scheduling strategies more scientifically to reduce network loss and improve energy utilization efficiency.

[0127] In this example, the income C from energy sales of the above-mentioned integrated energy service provider is Sell The corresponding expression can be as follows:

[0128]

[0129] Wherein, Md represents the number of the dth typical day in a year; d=1...D; D represents the total number of typical days; It represents the electric load power sold by the integrated energy service provider to the user side at time t; t = 1...T; T represents the total number of hours, for example, T can be 24; It represents the price of electricity sold by the integrated energy service provider to the user at time t; It represents the heat load power sold by the integrated energy service provider to the user at time t; It represents the heat price sold by the integrated energy service provider to the user at time t; It represents the cooling load power sold by the integrated energy service provider to the user at time t; It represents the price of cooling sold by the integrated energy service provider to the user at time t. In this example, by breaking down the energy sales revenue to each moment of each typical day and each energy type (electricity, heat, and cooling), the revenue obtained by the integrated energy service provider through energy sales during operation can be measured more accurately. Accurate calculation of energy sales revenue is conducive to comprehensive consideration of the complementarity of multiple energy types such as electricity, heat, and cooling, and more scientific formulation of energy production and supply strategies through multi-energy complementarity and coordinated optimization, so as to meet the needs of the user side and maximize its own revenue.

[0130] In this example, the energy purchase cost C of the above integrated energy service provider is buy The corresponding expression can be as follows:

[0131]

[0132] Among them, M d represents the number of the dth typical day in a year; d = 1...D; D represents the total number of typical days; represents the electric energy purchased by the integrated energy service provider from the distribution company at time t on the dth typical day; represents the price of electricity purchased by the integrated energy service provider from the distribution company at time t on the dth typical day; represents the electric energy purchased by the integrated energy service provider from the distributed generation generator on the dth typical day; represents the price of electric energy purchased by the integrated energy service provider from the distributed power generation company on the dth typical day; t=1...T; T represents the total number of hours, for example, T can be 24; represents the fuel cost of the integrated energy service provider at time t on the dth typical day; Δt represents the time interval; in this example, by breaking down the energy purchase cost to each moment of each typical day and each energy source (distribution company, distributed power generator), the cost of purchasing energy by the integrated energy service provider during operation can be measured more accurately. By accurately calculating the energy purchase cost, the integrated energy service provider can compare the prices of different energy suppliers and select energy suppliers with lower prices and reliable quality, thereby reducing procurement costs. While optimizing the energy procurement strategy, the integrated energy service provider can comprehensively consider the production, storage, conversion and consumption of energy, and improve the overall efficiency and reliability of the energy system through collaborative optimization.

[0133] In this example, based on the variable operating efficiency characteristics of the internal combustion generator and boiler, the relationship between their output power and fuel cost is expressed in the form of a quadratic function, that is, the fuel cost of the above-mentioned integrated energy service provider at time t on the dth typical day is The corresponding expression can be as follows:

[0134]

[0135] Among them, a e represents a first cost coefficient of the internal combustion generator; represents the output power of the internal combustion generator at time t; b e represents the second cost coefficient of the internal combustion generator; c e represents the third cost coefficient of the internal combustion generator; a h represents the first cost coefficient of the boiler; represents the output thermal power of the boiler at time t; b h represents the second cost coefficient of the boiler; c h Represents the third cost coefficient of the boiler; if for clean energy such as wind and light, and auxiliary equipment such as waste heat recovery and heat exchanger, its fuel cost can be considered to be 0; in this example, the variable operating efficiency characteristics of the internal combustion generator and the boiler are taken into account, and the nonlinear relationship between the output power and the fuel cost is reflected in the form of a quadratic function, and it is clearly pointed out that for clean energy such as wind and light, and auxiliary equipment such as waste heat recovery and heat exchanger, its fuel cost is 0, which is conducive to comprehensive energy service providers to give priority to the use of clean energy and the energy-saving effect of auxiliary equipment when formulating energy conversion strategies, so as to reduce dependence on traditional fuels.

[0136] Integrated energy service providers are service providers that use the technology and infrastructure provided by the integrated energy system to provide customers with comprehensive energy solutions. The integrated energy system includes a variety of distributed power generation systems and structures. Its investment cost is related to the capacity of the designed system. For example, the above investment cost C Cap The corresponding expression can be as follows:

[0137]

[0138] Among them, p inv，e represents the investment cost per unit capacity of equipment e in the integrated energy system; S equ，e represents the installed capacity of equipment e in the integrated energy system; N IESRepresents the number of devices in the integrated energy system. In this example, by breaking down the investment cost into each device and considering its unit capacity investment cost and installation capacity, the total investment cost of the integrated energy system can be evaluated more accurately. Accurate investment cost evaluation is conducive to integrated energy service providers making more reasonable investment decisions, optimizing system configuration and reducing investment risks when designing and planning integrated energy systems.

[0139] In this example, the above operation and maintenance cost C OM It can include personnel management and equipment maintenance costs, and the corresponding expression can be as follows:

[0140]

[0141] in, P represents the maintenance cost per unit power of equipment e in the integrated energy system, that is, the maintenance equipment cost; IES，e (t) represents the operating power of device e in the integrated energy system at time t; in this example, when determining the operation and maintenance costs, the integrated energy service provider can comprehensively consider the performance, operating status and maintenance requirements of the equipment, formulate targeted maintenance strategies, and improve the reliability and stability of the equipment. By monitoring the operating power and maintenance costs of the equipment, the integrated energy service provider can promptly detect potential failures of the equipment and take preventive maintenance measures to avoid failures and reduce downtime and maintenance costs.

[0142] In this example, the environmental cost of the integrated energy service provider mentioned above is The corresponding expression can be as follows:

[0143]

[0144] Among them, M d represents the number of the dth typical day in a year; d = 1...D; D represents the total number of typical days; P represents the unit carbon trading price of the integrated energy service provider at time t on the dth typical day; t=1...T; T represents the total number of hours; P Gas，t，d represents the natural gas purchase power of the integrated energy service provider at time t on the dth typical day; α Gas Represents the carbon dioxide emission coefficient of natural gas. In this example, quantifying environmental costs is conducive to promoting energy transformation, shifting from traditional fossil energy to clean energy and renewable energy, and achieving sustainable development of the energy industry. By optimizing the energy structure and improving energy utilization efficiency, integrated energy service providers can reduce energy and environmental costs and improve economic benefits.

[0145] When constructing a multi-agent optimization loss reduction multi-objective model, the optimization objectives are defined by the above objective function. In order to make these objectives feasible in actual operations and to ensure the robustness of the constructed model, corresponding constraints need to be formulated for the objective function.

[0146] For example, the above constraints may include one or more of the following: power balance constraints, branch voltage constraints, upper grid power purchase constraints, real-time electricity price constraints, power constraints, power purchase and sale constraints, and distribution network power flow equation constraints.

[0147] In this example, the expression corresponding to the power balance constraint can be as follows:

[0148] P MT，t +P DG，t +P Grid，t =P Load，t +P Loss，t +P IES，t ;

[0149] Among them, P MT，t represents the power generation of the micro gas turbine in period t; P DG，t P represents the amount of electricity purchased by the distributed generation generator in period t; Grid，t P represents the amount of electricity purchased from the main grid during period t; Load，t P represents the amount of electricity purchased by users during period t; Loss，t Indicates the power loss during period t; P IES，t represents the amount of electricity purchased by the integrated energy power generator at time t; in this example, by ensuring the balance between the power generation of the micro gas turbine, the power purchased by the distributed power generator, the power purchased from the main grid and the power purchased by the user, the lost power, and the power purchased by the integrated energy power generator, the stable operation of the system can be maintained, power shortages or surpluses can be avoided, and under the premise of satisfying the power balance constraint, the system can supply power more reliably, reduce power outages or insufficient power supply caused by the imbalance between power supply and demand, and thus improve users' power satisfaction.

[0150] In this example, the branch voltage constraint can be expressed as follows:

[0151] U l.t.min ≤U l，t ≤U l.t.max ;

[0152] Among them, U l，t represents the voltage amplitude of branch l during period t; U l.t.min Indicates the lower limit of the voltage allowed for branch l during period t; U l.t.max It represents the upper limit of the voltage allowed for branch l during period t;

[0153] In this example, the expression corresponding to the power purchase constraint of the upper-level power grid can be as follows:

[0154]

[0155] Among them, P Grid，i represents the amount of electricity purchased by the upper power grid in node i; It represents the lower limit of the amount of electricity that can be provided by the upper power grid in node i; It represents the upper limit of the amount of electricity that the upper power grid can provide in node i;

[0156] In this example, the expression corresponding to the real-time electricity price constraint can be as follows:

[0157] C Sell.t.min ≤C Sell，t ≤C Sell.t.max ;

[0158] Among them, C Sell，t represents the real-time electricity price during period t; C Sell.t.min represents the minimum real-time electricity price during period t; C Sell.t.max Indicates the maximum value of the real-time electricity price during period t;

[0159] In this example, the expression corresponding to the power constraint can be as follows:

[0160]

[0161] in, It represents the output power of the electric boiler during period t; Indicates the minimum output power of the electric boiler during period t; Indicates the maximum output power of the electric boiler during period t; represents the output power of the internal combustion generator during period t; It represents the minimum output power of the internal combustion generator during period t; represents the maximum output power of the internal combustion generator during period t; P Gas，t represents the gas purchase amount during period t; Represents the upper limit of gas purchase in time period t; in this example, by setting the minimum and maximum output power constraints of the electric boiler and the internal combustion generator, it is possible to ensure that these devices operate within a safe and stable range, avoiding equipment damage or safety accidents caused by overload or insufficient power; by setting the upper limit of gas purchase, the stability of the gas supply system can be ensured, avoiding problems such as excessive system pressure or insufficient supply caused by excessive gas purchase. Therefore, formulating power constraints is conducive to balancing the energy supply and demand relationship of the system, ensuring that the system maintains the stability and economy of energy supply while meeting the electricity and heat needs.

[0162] In this example, the expression corresponding to the power purchase and sales constraint can be as follows:

[0163]

[0164] in, P represents the power of DC power purchased by the integrated energy service provider during period t; DG.b The power of electricity purchased by the integrated energy service provider; It represents the minimum value of the power purchased by the integrated energy service provider in period t; It represents the maximum value of the power purchased by the integrated energy service provider in period t; represents the power sold by the integrated energy service provider during period t; Represents the maximum value of the power sold by the integrated energy service provider in time period t; in this example, by setting the minimum and maximum values ​​of the purchased power, it can be ensured that the integrated energy service provider will not exceed its power demand and the power supply capacity of the power grid during the power purchase process, thereby ensuring the balance of power supply and demand; by setting the maximum value of the sold power, it can be ensured that the integrated energy service provider will not exceed its power generation capacity and the acceptance range of the power grid during the power sales process, further maintaining the stability and reliability of the power market. Therefore, through clear constraints on power purchase and sales, integrated energy service providers can participate more actively in power market transactions, optimize the allocation of power resources, and achieve fair, just and efficient operation of the power market.

[0165] In this example, the expression corresponding to the distribution network flow equality constraint can be as follows:

[0166]

[0167] Among them, P Gi represents the active power of the system power supply injected into node i; Q Gi represents the reactive power of the system power injected into node i; P DGi represents the active power of the distributed generation injected into node i; Q DGi represents the reactive power of the distributed generation injected into node i; P Li represents the active load injected into node i; Q Li represents the reactive load injected into node i; U i represents the node voltage injected into node i; U j represents the node voltage of the injection node i directly connected to the injection node i; j=1...n; n represents the total number of nodes; G ij represents the real part of the admittance matrix of injection node i and injection node j; B ij represents the imaginary part of the admittance matrix of injection node i and injection node j; δ ijrepresents the phase angle difference between injection node i and injection node j; in this example, the power flow equation constraint reflects the power balance relationship between nodes in the distribution network, which is an inevitable reflection of the actual operation state of the distribution network. By incorporating these equation constraints into the distribution network state estimation model, the rationality and accuracy of the state estimation can be significantly improved. In the observability analysis of the distribution network, the power flow equation constraint can help infer the information of the missing measurement nodes. When the measurement data of some nodes are missing, the power flow equation constraint and the measurement data of other known nodes can be used to calculate the state information such as power and voltage of these missing nodes, which enhances the observability of the distribution network and enables the system operator to better understand and control the operation state of the entire distribution network; in addition, the power flow equation constraint can also be used for the operation optimization and planning of the distribution network. In the operation process of the distribution network, various safety, reliability and economic constraints need to be met. As one of the important constraints, the power flow equation constraint can help the system operator formulate a more reasonable operation strategy and scheduling plan to ensure that the distribution network operates in a safe, stable and economical state.

[0168] Therefore, in the process of constructing the model, the present invention not only improves the feasibility, accuracy and robustness of the model by formulating reasonable constraints, but also ensures that the multi-agent optimization and loss reduction multi-objective model meets various restrictions and conditions in actual operations while pursuing the maximization goal, thereby finding a more practical and reliable solution.

[0169] According to the objective function and constraints obtained by the above method, a multi-agent optimization loss reduction multi-objective model can be constructed. The next step is to solve the multi-agent optimization loss reduction multi-objective model. In order to ensure the effectiveness and practicality of the model, it is necessary to select appropriate optimization algorithms and solution methods to find the optimal or approximately optimal solution. Specifically:

[0170] In one implementation, Figure 2 As shown, the above-mentioned process of solving the pre-built multi-agent optimization loss reduction multi-objective model based on the operation data using the harmony algorithm to obtain the distribution network operation plan may include:

[0171] Initializing a harmony memory library and harmony parameters, the harmony parameters including: a harmony memory library, a harmony memory considering rate (HMCR), and the number of iterations, each initial harmony in the harmony memory library is the operating data of the distribution network, and the fitness function of the harmony is represented by the objective function;

[0172] The initial harmony memory consideration rate is updated by using a dynamic adjustment calculation formula to obtain an updated harmony memory consideration rate; a new harmony is created based on each initial harmony according to the updated harmony memory consideration rate, and a fitness function value corresponding to the new harmony is calculated according to a fitness function, and a fast non-dominated sorting is performed on the fitness function values ​​corresponding to the new harmony;

[0173] Based on the fitness function value corresponding to the new harmony and the fitness function value corresponding to the initial harmony, the harmony memory is iteratively updated using the elite retention strategy until the number of iterations is reached; the harmony memory updated when the number of iterations is reached is used as the final harmony memory (i.e., the Pareto solution set is obtained);

[0174] By using the intuitionistic fuzzy theory and the approximate ideal solution sorting method, the closeness evaluation value of each harmony in the final harmony memory library is calculated, and the harmony with the highest closeness evaluation value is selected as the distribution network operation plan; in this implementation method, in view of the various uncertain factors in the operation of the distribution network, such as load fluctuations, changes in the output of distributed power sources, etc., the harmony memory library is iteratively updated by the harmony algorithm, and the organic combination of global search and local optimization is realized. During the search process, the algorithm can continuously explore new solution spaces, and at the same time use the elite retention strategy to retain excellent solutions, thereby gradually approaching the optimal solution; and by dynamically adjusting the harmony memory consideration rate during the search process, the ability of global search and local search can be balanced during the search process, avoiding falling into the local optimal solution too early; at the same time, the fitness of the harmony is evaluated by using fast non-dominated sorting, which is beneficial to improving the search efficiency and accuracy.

[0175] For example, the above dynamic adjustment calculation formula can be as follows:

[0176]

[0177] Where HMCR(t0) represents the harmonic memory consideration rate at the t0th iteration; HMCR Min Indicates the preset minimum threshold of harmony memory consideration rate; HMCR Max Indicates the preset maximum threshold of harmony memory consideration rate; T Max Represents the maximum number of iterations; in this example, as the number of iterations increases, the harmonic memory consideration rate gradually increases from the minimum value to the maximum value. This dynamic adjustment is conducive to the algorithm to conduct a wider global search in the early stage to explore new solution spaces. As the iterations proceed, the algorithm gradually turns to local search to conduct a fine search near the excellent solutions that have been discovered, thereby accelerating convergence to the optimal solution and avoiding premature convergence to improve search efficiency.

[0178] In one implementation, the process of creating a new harmony based on each initial harmony according to the updated harmony memory consideration rate may include:

[0179] If the updated harmony memory consideration rate is greater than a randomly generated random number, a harmony is randomly selected from the harmony memory library as the created new harmony;

[0180] If the updated harmony memory consideration rate is not greater than the random number, a harmony is randomly generated from the preset solution space as the created new harmony; in this implementation, according to the comparison result of the updated HMCR and the random number, the algorithm can adaptively select the search strategy. When the HMCR is higher, the algorithm is more inclined to search in the harmony memory library to utilize the excellent solutions that have been discovered. When the HMCR is lower, the algorithm is more inclined to search in the preset solution space to explore new solution spaces. This adaptive search strategy is conducive to improving search efficiency, increasing search diversity, and helping the algorithm to jump out of the local optimal solution; in practical applications, the operating data of the distribution network may be uncertain or fluctuating. By introducing the comparison of random numbers, the algorithm can avoid repeatedly selecting the same harmony in the harmony memory library, which reduces redundancy and repetition in the search process and improves search efficiency.

[0181] In one implementation, the process of updating the harmony memory library based on the fitness function value corresponding to the new harmony and the fitness function value corresponding to the initial harmony by using the elite retention strategy may include:

[0182] The harmony with the smallest fitness function value among the initial harmony is taken as the underlying harmony;

[0183] The fitness function value of the underlying harmony is compared with the fitness function value corresponding to the new harmony. If the fitness function value of the underlying harmony is less than the fitness function value corresponding to the new harmony, the new harmony is used as the new underlying harmony, and the harmony memory is updated according to the new underlying harmony. In this implementation, the elite retention strategy ensures that only the harmony with the fitness function value of the underlying harmony (i.e., the harmony with the lowest fitness function value) in the harmony memory is retained, which is conducive to the harmony memory to continue to contain excellent solutions, thereby improving the quality of the entire search process. As the number of iterations increases, new harmonies are continuously generated and evaluated. Through the elite retention strategy, poor solutions are gradually eliminated, and excellent solutions are retained and used for subsequent searches. This mechanism promotes continuous optimization of the solution and is conducive to the algorithm eventually converging to a better solution.

[0184] The final harmony memory library (i.e., the Pareto solution set) obtained by the above-mentioned harmony algorithm is diverse, so even if the Pareto solution set is obtained, the optimal solution cannot be determined. In order to solve this problem, the present invention can determine the weight based on the expert scoring method, and consider the fuzziness and uncertainty of information in the decision-making process. The intuitive fuzzy theory is combined with the Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) to optimize the solution. Specifically:

[0185] In one implementation, the process of calculating the closeness evaluation value of each harmony in the final harmony memory library by using the intuitive fuzzy theory and the approximate ideal solution sorting method may include:

[0186] According to the final harmony memory library, the decision matrix is ​​established by using the intuitive fuzzy theory method, and the membership, non-membership and hesitation of the decision matrix are calculated. The decision matrices involved here are all standardized decision matrices.

[0187] Based on the membership of the decision matrix (for example, u h ), non-membership (for example, π h ) and hesitation (for example, v h Represented), the approximate ideal solution sorting method is used to calculate the closeness evaluation value of each harmony in the final harmony memory library; in this implementation, the intuitive fuzzy theory can more comprehensively describe the fuzziness of the decision object by introducing three parameters: membership, non-membership and hesitation, which is conducive to more accurately reflecting the actual situation and improving the accuracy of decision-making when dealing with complex and uncertain decision-making problems; by standardizing the decision matrix, the dimensional differences between different indicators can be eliminated, so that each indicator is comparable and consistent in the decision-making process, which is conducive to improving the reliability and accuracy of the decision; the closeness evaluation value is an important indicator to measure the quality of each harmony. In this implementation, by calculating the closeness evaluation value, the differences between different harmonies can be intuitively compared, providing decision makers with a clear decision basis; at the same time, the closeness evaluation value can also be flexibly adjusted according to actual needs to adapt to different decision-making scenarios. The TOPSIS method can comprehensively evaluate the pros and cons of each harmony by calculating the closeness of each harmony to the ideal solution and the negative ideal solution. This method not only considers the similarity between the harmony and the ideal solution, but also considers the difference between the harmony and the negative ideal solution, thereby enhancing the comprehensiveness of the decision.

[0188] For example, the above membership u h The corresponding expression can be expressed as:

[0189] For example, the above non-membership degree π h The corresponding expression can be expressed as: h =1-u h -v h ;

[0190] For example, the above hesitation v h The corresponding expression can be expressed as:

[0191] Among them, a h1 Indicates the indicator satisfaction value; b h1 Indicates the unsatisfactory value of the indicator; c h1 represents the indicator hesitation value;

[0192] In this implementation, the process of calculating the closeness evaluation value of each harmony in the final harmony memory library by using the approximate ideal solution sorting method based on the membership, non-membership and hesitation of the decision matrix may include:

[0193] According to the membership, non-membership and hesitation of the decision matrix, the expert scoring method is used to determine the weight of the indicators, and the distance between each harmony in the final harmony memory library and the preset theoretical optimal solution and theoretical worst solution is calculated using the approximate ideal solution sorting method (for example, the distance between each harmony and the theoretical optimal solution is calculated using The distance between each harmony and the theoretical worst solution is expressed as express);

[0194] According to the distance between each harmony and the theoretical optimal solution and the theoretical worst solution, the closeness evaluation value corresponding to each harmony in the final harmony memory library is calculated (for example, M h In this implementation, by calculating the distance between each harmony and the theoretical optimal solution and the theoretical worst solution, as well as the closeness evaluation value, the quality of each harmony can be intuitively displayed, and the differences between different harmonies can be compared, so as to find excellent harmony more quickly, which is conducive to the harmony algorithm to converge to the optimal solution or approximate optimal solution more quickly during the search process, thereby improving the search efficiency; and by comprehensively considering multiple dimensions and introducing the expert scoring method to determine the weights, the quality of each harmony can be more accurately evaluated, which is conducive to the harmony algorithm to maintain robustness when dealing with complex and changeable decision-making problems, thereby enhancing the robustness of the algorithm.

[0195] For example, the distance between each harmony and the theoretical optimal solution The corresponding expression can be:

[0196]

[0197] For example, the distance between each harmony and the theoretical worst solution The corresponding expression can be:

[0198]

[0199] For example, the closeness evaluation value M h The corresponding expression can be:

[0200]

[0201] Among them, w p′ represents the weight of the p′th indicator determined by the expert scoring method; p′=1…s; s represents the total number of indicators; Indicates the degree of membership between harmony and theoretical optimal solution; u hp′ represents the p′th index membership; represents the non-membership degree between harmony and theoretical optimal solution; v hp′ represents the non-membership degree of the p′th index; Represents the hesitation between harmony and theoretical optimal solution; π hp′ represents the hesitation of the p′th indicator; Indicates the degree of membership between harmony and the theoretical worst solution; Indicates the non-membership degree between harmony and the theoretically worst solution; Indicates the hesitation between harmony and theoretical worst solution;

[0202] In summary, the prior art only performs loss reduction optimization on a single subject without considering the influence of multiple subjects in the market, and the prior art is prone to fall into local optimality and slow convergence when using traditional optimization algorithms to solve multi-objective models. The present invention proposes a multi-subject collaborative loss reduction multi-objective optimization method based on a harmony algorithm, and establishes a multi-objective model by considering the economic, technical, environmental protection and other aspects of different market subjects. The parameters required in the model are easy to obtain in reality and have good usability. It can not only reflect the interests of different subjects, but also reflect the influence of the collaborative operation of multiple subjects in the market on the loss reduction optimization of the distribution network. The improved harmony algorithm is used to solve the multi-subject optimization loss reduction multi-objective model. The solution process is simple and easy to expand, does not rely on expensive foreign commercial solvers, can reduce the cost of use, and the optimal decision made takes into account the economic benefits of each market subject, and has the function of reducing the system line loss as a whole.

[0203] Embodiment 2:

[0204] As an intelligent optimization algorithm, the harmony algorithm has been widely used due to its good optimization ability and robustness, but it has the problem of being easily trapped in local optimality and slow convergence speed. To address this problem, an embodiment is used to specifically illustrate the process of using the harmony algorithm in the present invention to solve the multi-agent optimization and loss reduction multi-objective model. In this embodiment, the traditional harmony algorithm is improved, and dynamic adjustment parameters (that is, the harmony memory consideration rate is updated through the above-mentioned dynamic adjustment calculation formula) and the elite retention strategy are introduced in the process to achieve the solution of the multi-agent optimization and loss reduction multi-objective model. In the optimization process using the improved harmony algorithm, the decision variables are represents the j1th tone corresponding to the island x, and n1 represents the total number of tones that make up a harmony) as the tone of an instrument (specifically, it may include: the installed capacity S of the device e in the integrated energy system equ，e , the output power of the internal combustion generator at time t The output thermal power of the boiler at time t The operating power P of equipment e in the integrated energy system at time t IES，e (t), the electric load power sold by the comprehensive energy service provider to the user at time t The heat load power sold by the integrated energy service provider to the user at time t The cooling load power sold by the integrated energy service provider to the user at time t and the natural gas purchase power PGas,t,d) of the integrated energy service provider at time t on the dth typical day, and the distribution operation plan (also known as the island combination plan) is regarded as a harmony composed of the tone j1 (harmony by can be interpreted as the j1th tone corresponding to island 1, and so on. can be interpreted as the j1th tone corresponding to the island n1), and the population composed of different harmonies {P 1 , P 2 ,…,P HMS} as Harmony Memory (HM); where P HMS Indicates the HMSth harmony. Specifically, the process of solving the multi-agent optimization loss reduction multi-objective model by the harmony algorithm used in this embodiment is as follows:

[0205] (1) Generation of harmony memory HM: HMS harmonies are randomly generated and put into the harmony memory to store a certain number of harmonies. Each harmony represents a possible distribution network operation plan, including the output of distributed generators, load distribution, and scheduling strategies of energy service providers. According to the characteristics of the optimization problem of the present invention, the harmony memory is used to store a certain number of harmonies. represents the load point j1 contained in the island x, that is, the j1th tone corresponding to the island x in, It can be interpreted as the j1th tone corresponding to the load point x1 in the island x, and so on. can be interpreted as the load point in the island x The corresponding j1th tone.

[0206] (2) Initialize the harmony memory HM

[0207] HMS harmonies are randomly generated, each of which contains all decision variables (such as the output power of distributed generators, load distribution ratio, energy trading strategy of energy service providers, etc.).

[0208] (3) Dynamically adjust parameters

[0209] As the number of iterations increases, the value of the harmony memory consideration rate HMCR can be gradually reduced to increase the proportion of randomly generated parts in new harmonies and enhance the global search capability of the algorithm.

[0210] (4) Create new harmony: Create new harmony In order to make full use of the information accumulated by HM and to appropriately randomly extract values ​​from outside HM, the following expression is used to create a new harmony:

[0211]

[0212] Where: M x is the feasible domain of decision variables; HMCR represents the harmonic memory consideration rate; A new harmony is created; rand represents a randomly generated random number; a collection Indicates the harmony stored in the harmony memory (e.g. Indicates the first harmony corresponding to island x in the harmony memory library, and so on. represents the HMSth harmony corresponding to island x in the harmony memory bank).

[0213] (4) Evaluate the new harmony: Calculate the fitness function value corresponding to the new harmony.

[0214] (5) Update the harmony memory and terminate the iteration.

[0215] If the fitness function value of the new harmony is better than the fitness function value of the worst harmony in the harmony memory, the elite preservation strategy is implemented to replace it. After each iteration, the number of harmonies in the harmony memory remains unchanged to ensure that the optimal solution is retained in subsequent iterations. When the preset number of iterations is reached, the algorithm terminates and outputs the optimal harmony in the harmony memory as the solution to the problem. The elite retention strategy is conducive to improving the convergence speed of the algorithm and the accuracy of the optimization results. The elite retention strategy can avoid losing excellent individuals and retain the harmonies with high objective function values ​​in the parent harmony memory as elite individuals to the next generation.

[0216] In this embodiment, by specifically describing the process of solving the model by the harmony algorithm, it can be explained that the harmony algorithm adopted by the present invention can increase the proportion of randomly generated parts in the new harmony by dynamically adjusting the harmony memory consideration rate and gradually reducing the value of the harmony memory consideration rate as the number of iterations increases. This is beneficial for the harmony algorithm to jump out of the local optimal solution during the search process and enhance the global search capability, thereby finding a better solution; and by introducing the elite retention strategy, the optimal harmony in the harmony memory library is retained after each iteration, and the worst harmony is replaced, which ensures that the optimal solution is retained in subsequent iterations, avoids the blindness of the algorithm in the search process, and thus improves the convergence speed and the accuracy of the optimization results.

[0217] Embodiment 3:

[0218] The present invention based on the same inventive concept also provides a multi-agent collaborative loss reduction multi-objective optimization system based on a harmony algorithm, the structural composition diagram is as follows Figure 3 As shown, including:

[0219] A data acquisition module, used to acquire the operation data of the distribution network;

[0220] The model solving module is used to solve the pre-built multi-agent optimization loss reduction multi-objective model based on the operation data using the harmony algorithm to obtain the distribution network operation plan;

[0221] The collaborative loss reduction module is used to collaboratively reduce losses for each revenue entity in the distribution network according to the distribution network operation plan;

[0222] Among them, the multi-agent optimization loss reduction multi-objective model is constructed with the goals of maximizing the comprehensive net income of each benefit entity, the user's energy purchase satisfaction and the new energy consumption. Each benefit entity includes at least several of the following: distribution companies, integrated energy service providers, and distributed power generation companies.

[0223] In one implementation, the optimization system may further include: a model building module, which may be used to:

[0224] Calculate the comprehensive net income of each income subject based on the income from energy sales and the total operating costs of each income subject;

[0225] Calculate the user's energy purchase satisfaction based on the user's purchased energy amount and the preference coefficient corresponding to the purchased energy amount;

[0226] Calculate the new energy consumption rate based on the new energy consumption and actual power generation;

[0227] The objective function is constructed with the goal of maximizing the comprehensive net income of each income subject, the user's energy purchase satisfaction and the new energy consumption rate;

[0228] According to the objective function, formulate the constraint conditions corresponding to the objective function;

[0229] Based on the objective function and constraints, a multi-objective model of multi-agent optimization and loss reduction is constructed.

[0230] For example, the total operating cost of the above-mentioned distribution company may include one or more of the following: the distribution company's energy purchase cost, operation and maintenance cost, the distribution company's electricity purchase cost from the superior main grid, the distribution company's electricity purchase cost from distributed power generators, the distribution company's environmental cost and network loss cost;

[0231] For example, the total operating cost of the above-mentioned integrated energy service provider may include one or more of the following: energy purchase cost of the integrated energy service provider, investment cost, environmental cost of the integrated energy service provider, operation and maintenance cost, and network loss cost;

[0232] For example, the overall income from energy sales of the above-mentioned distributed power generation companies may include one or more of the following: income from energy sales of distributed power generation companies and income from photovoltaic grid-connected subsidies;

[0233] For example, the total operating cost of the above-mentioned distributed power generation company may include one or more of the following: unit operation and maintenance cost and unit power generation fuel cost;

[0234] For example, the above-mentioned purchased energy amount may include one or more of the following: purchased heat amount and purchased electricity amount;

[0235] For example, the above-mentioned actual power generation may include one or more of the following: actual photovoltaic power generation and actual wind power generation;

[0236] For example, the above constraints may include one or more of the following: power balance constraints, branch voltage constraints, upper grid power purchase constraints, real-time electricity price constraints, power constraints, power purchase and sale constraints, and distribution network power flow equation constraints.

[0237] For example, the expression corresponding to the above objective function can be as follows:

[0238]

[0239] Among them, B DC represents the comprehensive net income of the distribution company; F IES F represents the comprehensive net income of the integrated energy service provider; DG F represents the comprehensive net income of distributed power generators; user Indicates the user's purchasing satisfaction; F RES Indicates the amount of new energy consumed; B sell Represents the revenue from selling electricity to distribution companies; C In represents the energy purchase cost of the distribution company; C OM represents the operation and maintenance cost; C Buy.grid C represents the cost of electricity purchased by the distribution company from the upper grid; buy.DG E represents the cost of electricity purchased by the distribution company from the distributed generation generator; DC represents the environmental cost of the distribution company; C Loss represents the network loss cost; C Sell Represents the income from energy sales of integrated energy service providers; C buy represents the energy purchase cost of the integrated energy service provider; C Cap represents the investment cost; represents the environmental cost of the integrated energy service provider; R Sell Represents the revenue of the distribution company from energy sales; R Gusb Represents the income from photovoltaic grid-connected subsidies; C Oper represents the unit operation and maintenance cost of the distributed generation generator; C Cgen represents the unit power generation fuel cost of distributed generation generators; v e Indicates the first preference coefficient corresponding to the amount of electricity purchased by the user; represents the amount of electricity purchased by the user during period t; αe represents the second preference coefficient corresponding to the amount of electricity purchased by the user; a′ represents the constant coefficient; v h Indicates the first preference coefficient corresponding to the user's heat purchase; α h Indicates the second preference coefficient corresponding to the user's purchase of heat; Indicates the amount of heat purchased by users during period t; represents the actual photovoltaic power generation in the hth hour; represents the actual wind power generation in hour h; represents the photovoltaic power generation purchased by the integrated energy service provider from the distributed power generation company at hour h; It represents the amount of wind power purchased by the integrated energy service provider from the distributed generation power generator at the hth hour.

[0240] In one implementation, the above-mentioned model solving module may include:

[0241] The parameter initialization submodule is used to initialize the harmony memory library and harmony parameters. The harmony parameters include: harmony memory library, harmony memory consideration rate, and number of iterations. Each initial harmony in the harmony memory library is the operating data of the distribution network, and the fitness function of the harmony is represented by the objective function.

[0242] The harmony creation submodule is used to update the initial harmony memory consideration rate by using a dynamic adjustment calculation formula to obtain an updated harmony memory consideration rate; create a new harmony based on each initial harmony according to the updated harmony memory consideration rate, and calculate the fitness function value corresponding to the new harmony according to the fitness function;

[0243] A memory library update submodule is used to iteratively update the harmony memory library based on the fitness function value corresponding to the new harmony and the fitness function value corresponding to the initial harmony, using an elite retention strategy, until the number of iterations is reached; and the harmony memory library updated when the number of iterations is reached is used as the final harmony memory library;

[0244] The closeness evaluation submodule is used to calculate the closeness evaluation value of each harmony in the final harmony memory library by using the intuitive fuzzy theory and the approximate ideal solution sorting method, and select the harmony with the highest closeness evaluation value as the distribution network operation plan;

[0245] For example, the operation data of the above-mentioned distribution network may include one or more of the following: output power of distributed generators, load distribution ratio, and energy trading strategy of energy service providers.

[0246] For example, the above dynamic adjustment calculation formula can be as follows:

[0247]

[0248] Where HMCR(t0) represents the harmonic memory consideration rate at the t0th iteration; HMCR Min Indicates the preset minimum threshold of harmony memory consideration rate; HMCR Max Indicates the preset maximum threshold of harmony memory consideration rate; T Max Indicates the maximum number of iterations.

[0249] In one implementation, the harmony creation submodule may include:

[0250] An initial selection unit, used for randomly selecting a harmony from the harmony memory library as a newly created harmony when the updated harmony memory consideration rate is greater than a randomly generated random number;

[0251] The space generation unit is used to randomly generate a harmony from a preset solution space as a created new harmony when the updated harmony memory consideration rate is not greater than the random number.

[0252] In this implementation, the memory library update submodule may include:

[0253] A bottom layer setting unit, used for taking the harmony with the smallest fitness function value among the initial harmony as the bottom layer harmony;

[0254] The bottom-level replacement unit is used to compare the fitness function value of the bottom-level harmony with the fitness function value corresponding to the new harmony. If the fitness function value of the bottom-level harmony is less than the fitness function value corresponding to the new harmony, the new harmony is used as the new bottom-level harmony, and the harmony memory bank is updated according to the new bottom-level harmony.

[0255] In this implementation, the above-mentioned closeness evaluation submodule may include:

[0256] A matrix generation unit is used to establish a decision matrix based on the final harmony memory library using an intuitive fuzzy theory method, and calculate the membership, non-membership and hesitation of the decision matrix;

[0257] The closeness calculation unit is used to calculate the closeness evaluation value of each harmony in the final harmony memory library based on the membership, non-membership and hesitation of the decision matrix and using the approximate ideal solution sorting method.

[0258] In this implementation, the proximity calculation unit may include:

[0259] The distance calculation subunit is used to calculate the distance between each harmony in the final harmony memory bank and the preset theoretical optimal solution and theoretical worst solution respectively according to the membership, non-membership and hesitation of the decision matrix and by using the approximate ideal solution sorting method;

[0260] The evaluation unit is used to calculate the closeness evaluation value corresponding to each harmony in the final harmony memory library according to the distance between each harmony and the theoretical optimal solution and the theoretical worst solution.

[0261] Embodiment 4:

[0262] like Figure 4 As shown, the present invention also provides an electronic device, which may be a computer device, a single-chip device, an intelligent mobile device, etc. The electronic device in this embodiment may include a processor, a memory, a transceiver component, etc. The memory, the processor, and the transceiver component are connected via a bus; the memory may be used to store an execution program, and an exemplary execution program may include instructions; the processor is used to execute the instructions stored in the memory. The memory may also be used to store data, which may be called and / or modified when the instructions are executed.

[0263] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, which is suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions in a storage medium to implement the corresponding method flow or corresponding functions, so as to realize the steps of a multi-agent collaborative loss reduction and multi-objective optimization method based on a harmony algorithm in the above-mentioned embodiment.

[0264] Embodiment 5:

[0265] Based on the same inventive concept, the present invention also provides a readable storage medium, specifically an electronic device readable storage medium (Memory), which is a memory device in an electronic device for storing programs and data. It can be understood that the storage medium here can include both built-in storage media in electronic devices and, of course, extended storage media supported by electronic devices. The storage medium provides a storage space that stores the operating system of the terminal. In addition, one or more instructions suitable for being loaded and executed by a processor are also stored in the storage space, and these instructions can be one or more execution programs (including program codes). It should be noted that the storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk storage. The processor loads and executes one or more instructions stored in the storage medium, which can implement the steps of a multi-agent collaborative loss reduction multi-objective optimization method based on a harmony algorithm in the above embodiment.

[0266] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0267] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0268] These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0269] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0270] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit its protection scope. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the field should understand that after reading the present invention, those skilled in the art can still make various changes, modifications or equivalent substitutions to the specific implementation methods of the application, but these changes, modifications or equivalent substitutions are all within the protection scope of the claims to be approved.

Claims

1. A multi-agent collaborative loss reduction multi-objective optimization method based on harmony algorithm, characterized in that: include: Obtaining the operation data of the distribution network; Based on the operation data, a pre-built multi-agent optimization loss reduction multi-objective model is solved using a harmony algorithm to obtain a distribution network operation plan; Using the distribution network operation plan, collaboratively reducing losses for various revenue entities in the distribution network; Among them, the multi-agent optimization loss reduction multi-objective model is constructed with the goals of maximizing the comprehensive net income of each benefit entity, the user's energy purchase satisfaction and the new energy consumption rate. The benefit entities include at least several of the following: distribution companies, integrated energy service providers, and distributed power generation companies.

2. The method according to claim 1, characterized in that The multi-agent optimization loss reduction multi-objective model includes the following construction process: Calculate the comprehensive net income of each income subject based on the income from energy sales of each income subject and the total operating costs corresponding to each income subject; Calculating the user's energy purchase satisfaction according to the user's purchased energy amount and the preference coefficient corresponding to the purchased energy amount; Calculate the new energy consumption rate based on the new energy consumption and actual power generation; The objective function is constructed with the goal of maximizing the comprehensive net income of each income subject, the user's energy purchase satisfaction and the new energy consumption rate; According to the objective function, formulate constraint conditions corresponding to the objective function; Based on the objective function and the constraints, a multi-agent optimization and loss reduction multi-objective model is constructed.

3. The method according to claim 2, characterized in that The total operating cost of the distribution company includes one or more of the following: the distribution company's energy purchase cost, operation and maintenance cost, the distribution company's electricity purchase cost from the superior main grid, the distribution company's electricity purchase cost from distributed power generators, the distribution company's environmental cost and network loss cost; The total operating cost of the integrated energy service provider includes one or more of the following: energy purchase cost of the integrated energy service provider, investment cost, environmental cost of the integrated energy service provider, operation and maintenance cost, and network loss cost; The income from energy sales of the distributed power generation generator includes one or more of the following: income from energy sales of the distribution company and income from photovoltaic grid-connected subsidies; The total operating cost of the distributed power generator includes one or more of the following: unit operation and maintenance cost and unit power generation fuel cost; The amount of energy purchased includes one or more of the following: the amount of heat purchased and the amount of electricity purchased; The actual power generation includes one or more of the following: actual photovoltaic power generation and actual wind power generation; The constraint conditions include one or more of the following: power balance constraint, branch voltage constraint, upper power grid power purchase constraint, real-time electricity price constraint, power constraint, power purchase and sale constraint and distribution network power flow equation constraint.

4. The method according to claim 2 or 3, characterized in that The expression corresponding to the objective function is as follows: Among them, B DC represents the comprehensive net income of the distribution company; F IES F represents the comprehensive net income of the integrated energy service provider; DG F represents the comprehensive net income of distributed power generators; user Indicates the user's purchasing satisfaction; F RES represents the new energy consumption rate; B sell Represents the revenue from selling electricity to distribution companies; C In represents the energy purchase cost of the distribution company; C OM represents the operation and maintenance cost; C Buy.grid C represents the cost of electricity purchased by the distribution company from the upper grid; buy.DG E represents the cost of electricity purchased by the distribution company from the distributed generation generator; DC represents the environmental cost of the distribution company; C Loss represents the network loss cost; C Sell Represents the income from energy sales of integrated energy service providers; C buy represents the energy purchase cost of the integrated energy service provider; C Cap represents the investment cost; represents the environmental cost of the integrated energy service provider; R Sell Represents the revenue of the distribution company from energy sales; R Gusb Represents the income from photovoltaic grid-connected subsidies; C Oper represents the unit operation and maintenance cost of the distributed generation generator; C Cgen represents the unit power generation fuel cost of distributed generation generators; v e Indicates the first preference coefficient corresponding to the amount of electricity purchased by the user; represents the amount of electricity purchased by users during period t; α e represents the second preference coefficient corresponding to the amount of electricity purchased by the user; a′ represents the constant coefficient; v h Indicates the first preference coefficient corresponding to the user's heat purchase; α h Indicates the second preference coefficient corresponding to the user's purchase of heat; Indicates the amount of heat purchased by users during period t; represents the actual photovoltaic power generation in the hth hour; represents the actual wind power generation in hour h; represents the photovoltaic power generation purchased by the integrated energy service provider from the distributed power generation company at hour h; It represents the amount of wind power purchased by the integrated energy service provider from the distributed generation power generator at the hth hour.

5. The method according to claim 2, characterized in that Based on the operation data, the pre-built multi-agent optimization loss reduction multi-objective model is solved by using the harmony algorithm to obtain a distribution network operation plan, including: Initializing a harmony memory library and harmony parameters, wherein the harmony parameters include: a harmony memory library, a harmony memory consideration rate, and a number of iterations, wherein each initial harmony in the harmony memory library is the operating data of the distribution network, and the fitness function of the harmony is represented by the objective function; The initial harmony memory consideration rate is updated by using a dynamic adjustment calculation formula to obtain an updated harmony memory consideration rate; a new harmony is created based on each initial harmony according to the updated harmony memory consideration rate, and a fitness function value corresponding to the new harmony is calculated according to the fitness function; Based on the fitness function value corresponding to the new harmony and the fitness function value corresponding to the initial harmony, the harmony memory library is iteratively updated using an elite retention strategy until the number of iterations is reached; the harmony memory library updated when the number of iterations is reached is used as the final harmony memory library; Using intuitive fuzzy theory and the approximate ideal solution sorting method, the closeness evaluation value of each harmony in the final harmony memory library is calculated, and the harmony with the highest closeness evaluation value is selected as the distribution network operation plan; The operation data of the distribution network includes one or more of the following: output power of distributed generators, load distribution ratio and energy trading strategy of energy service providers.

6. The method according to claim 5, characterized in that The dynamic adjustment calculation formula is as follows: Where HMCR(t0) represents the harmony memory consideration rate at the t0th iteration; HMCR Min Indicates the preset minimum threshold of harmony memory consideration rate; HMCR Max Indicates the preset maximum threshold of harmony memory consideration rate; T Max Indicates the maximum number of iterations.

7. The method according to claim 5, characterized in that The method of creating a new harmony based on each initial harmony according to the updated harmony memory consideration rate includes: If the updated harmony memory consideration rate is greater than a randomly generated random number, randomly selecting a harmony from the harmony memory library as a newly created harmony; If the updated harmony memory consideration rate is not greater than the random number, a harmony is randomly generated from a preset solution space as a created new harmony.

8. The method according to claim 5, characterized in that Based on the fitness function value corresponding to the new harmony and the fitness function value corresponding to the initial harmony, the harmony memory library is updated using an elite retention strategy, including: The harmony with the smallest fitness function value among the initial harmony is used as the underlying harmony; The fitness function value of the underlying harmony is compared with the fitness function value corresponding to the new harmony. If the fitness function value of the underlying harmony is less than the fitness function value corresponding to the new harmony, the new harmony is used as a new underlying harmony, and the harmony memory bank is updated according to the new underlying harmony.

9. The method according to claim 5, characterized in that The method of using the intuitive fuzzy theory and the approximate ideal solution sorting method to calculate the closeness evaluation value of each harmony in the final harmony memory library includes: According to the final harmony memory library, a decision matrix is ​​established by using an intuitive fuzzy theory method, and the membership, non-membership and hesitation of the decision matrix are calculated; Based on the membership, non-membership and hesitation of the decision matrix, the closeness evaluation value of each harmony in the final harmony memory library is calculated using the approximate ideal solution sorting method.

10. The method according to claim 9, characterized in that The method of calculating the closeness evaluation value of each harmony in the final harmony memory library based on the membership, non-membership and hesitation of the decision matrix and using the approximate ideal solution sorting method includes: According to the membership, non-membership and hesitation of the decision matrix, the distance between each harmony in the final harmony memory bank and the preset theoretical optimal solution and theoretical worst solution is calculated by using the approximate ideal solution sorting method; According to the distances between each harmony and the theoretical optimal solution and the theoretical worst solution, respectively, a closeness evaluation value corresponding to each harmony in the final harmony memory library is calculated.

11. A multi-agent collaborative loss reduction multi-objective optimization system based on harmony algorithm, characterized in that: include: A data acquisition module, used to acquire the operation data of the distribution network; A model solving module, for solving a pre-built multi-agent optimization loss reduction multi-objective model based on the operation data using a harmony algorithm to obtain a distribution network operation plan; A collaborative loss reduction module, used to collaboratively reduce losses for each income subject in the distribution network according to the distribution network operation plan; Among them, the multi-agent optimization loss reduction multi-objective model is constructed with the goals of maximizing the comprehensive net income of each benefit entity, the user's energy purchase satisfaction and the new energy consumption rate. The benefit entities include at least several of the following: distribution companies, integrated energy service providers, and distributed power generation companies.

12. The system according to claim 11, characterized in that Also includes: Model building modules for: Calculate the comprehensive net income of each income subject based on the income from energy sales of each income subject and the total operating costs corresponding to each income subject; Calculating the user's energy purchase satisfaction according to the user's purchased energy amount and the preference coefficient corresponding to the purchased energy amount; Calculate the new energy consumption rate based on the new energy consumption and actual power generation; The objective function is constructed with the goal of maximizing the comprehensive net income of each income subject, the user's energy purchase satisfaction and the new energy consumption rate; According to the objective function, formulate constraint conditions corresponding to the objective function; Based on the objective function and the constraints, a multi-agent optimization and loss reduction multi-objective model is constructed.

13. The system of claim 12, wherein: The total operating cost of the distribution company includes one or more of the following: the distribution company's energy purchase cost, operation and maintenance cost, the distribution company's electricity purchase cost from the superior main grid, the distribution company's electricity purchase cost from distributed power generators, the distribution company's environmental cost and network loss cost; The total operating cost of the integrated energy service provider includes one or more of the following: energy purchase cost of the integrated energy service provider, investment cost, environmental cost of the integrated energy service provider, operation and maintenance cost, and network loss cost; The income from energy sales of the distributed power generation generator includes one or more of the following: income from energy sales of the distribution company and income from photovoltaic grid-connected subsidies; The total operating cost of the distributed power generator includes one or more of the following: unit operation and maintenance cost and unit power generation fuel cost; The amount of energy purchased includes one or more of the following: the amount of heat purchased and the amount of electricity purchased; The actual power generation includes one or more of the following: actual photovoltaic power generation and actual wind power generation; The constraint conditions include one or more of the following: power balance constraint, branch voltage constraint, upper power grid power purchase constraint, real-time electricity price constraint, power constraint, power purchase and sale constraint and distribution network power flow equation constraint.

14. The system of claim 13, wherein: The expression corresponding to the objective function is as follows: Among them, B DC represents the comprehensive net income of the distribution company; F IES F represents the comprehensive net income of the integrated energy service provider; DG F represents the comprehensive net income of distributed power generators; user Indicates the user's purchasing satisfaction; F RES represents the new energy consumption rate; B sell Represents the revenue from selling electricity to distribution companies; C In represents the energy purchase cost of the distribution company; C OM represents the operation and maintenance cost; C Buy.grid C represents the cost of electricity purchased by the distribution company from the upper grid; buy.DG E represents the cost of electricity purchased by the distribution company from the distributed generation generator; DC represents the environmental cost of the distribution company; C Loss represents the network loss cost; C Sell Represents the income from energy sales of integrated energy service providers; C buy represents the energy purchase cost of the integrated energy service provider; C Cap represents the investment cost; represents the environmental cost of the integrated energy service provider; R Sell Represents the revenue of the distribution company from energy sales; R Gusb Represents the income from photovoltaic grid-connected subsidies; C Oper represents the unit operation and maintenance cost of the distributed generation generator; C Cgen represents the unit power generation fuel cost of distributed generation generators; v e Indicates the first preference coefficient corresponding to the amount of electricity purchased by the user; represents the amount of electricity purchased by users during period t; α e represents the second preference coefficient corresponding to the amount of electricity purchased by the user; a′ represents the constant coefficient; v h Indicates the first preference coefficient corresponding to the user's heat purchase; α h Indicates the second preference coefficient corresponding to the user's purchase of heat; Indicates the amount of heat purchased by users during period t; represents the actual photovoltaic power generation in the hth hour; represents the actual wind power generation in hour h; represents the photovoltaic power generation purchased by the integrated energy service provider from the distributed power generation company at hour h; It represents the amount of wind power purchased by the integrated energy service provider from the distributed generation power generator at the hth hour.

15. The system of claim 12, wherein: The model solving module comprises: A parameter initialization submodule is used to initialize the harmony memory library and harmony parameters, wherein the harmony parameters include: harmony memory library, harmony memory consideration rate, and number of iterations, wherein each initial harmony in the harmony memory library is the operating data of the distribution network, and the fitness function of the harmony is represented by the objective function; A harmony creation submodule is used to update the initial harmony memory consideration rate using a dynamic adjustment calculation formula to obtain an updated harmony memory consideration rate; create a new harmony based on each initial harmony according to the updated harmony memory consideration rate, and calculate the fitness function value corresponding to the new harmony according to the fitness function; A memory library updating submodule, for iteratively updating the harmony memory library based on the fitness function value corresponding to the new harmony and the fitness function value corresponding to the initial harmony, using an elite retention strategy, until the number of iterations is reached; and using the harmony memory library updated when the number of iterations is reached as the final harmony memory library; A closeness evaluation submodule is used to calculate the closeness evaluation value of each harmony in the final harmony memory library by using intuitive fuzzy theory and approximate ideal solution sorting method, and select the harmony with the highest closeness evaluation value as the distribution network operation plan; The operation data of the distribution network includes one or more of the following: output power of distributed generators, load distribution ratio and energy trading strategy of energy service providers.

16. The system of claim 15, wherein: The dynamic adjustment calculation formula is as follows: Where HMCR(t0) represents the harmony memory consideration rate at the t0th iteration; HMCR Min Indicates the preset minimum threshold of harmony memory consideration rate; HMCR Max Indicates the preset maximum threshold of harmony memory consideration rate; T Max Indicates the maximum number of iterations.

17. The system of claim 15, wherein: The harmony creation submodule includes: an initial selection unit, configured to randomly select a harmony from the harmony memory library as a newly created harmony when the updated harmony memory consideration rate is greater than a randomly generated random number; The space generation unit is used to randomly generate a harmony from a preset solution space as a created new harmony when the updated harmony memory consideration rate is not greater than the random number.

18. The system of claim 15, wherein: The memory update submodule includes: A bottom layer setting unit, used for taking the harmony with the smallest fitness function value among the initial harmony as the bottom layer harmony; The bottom-level replacement unit is used to compare the fitness function value of the bottom-level harmony with the fitness function value corresponding to the new harmony. If the fitness function value of the bottom-level harmony is less than the fitness function value corresponding to the new harmony, the new harmony is used as a new bottom-level harmony, and the harmony memory bank is updated according to the new bottom-level harmony.

19. The system of claim 15, wherein: The closeness evaluation submodule includes: A matrix generating unit, used for establishing a decision matrix according to the final harmony memory library by using an intuitive fuzzy theory method, and calculating the membership, non-membership and hesitation of the decision matrix; The closeness calculation unit is used to calculate the closeness evaluation value of each harmony in the final harmony memory library based on the membership, non-membership and hesitation of the decision matrix and using the approximate ideal solution sorting method.

20. The system of claim 19, wherein: The closeness calculation unit includes: A distance calculation subunit, for calculating the distance between each harmony in the final harmony memory bank and the preset theoretical optimal solution and theoretical worst solution respectively, according to the membership, non-membership and hesitation of the decision matrix, using an approximate ideal solution sorting method; The evaluation unit is used to calculate the closeness evaluation value corresponding to each harmony in the final harmony memory library according to the distance between each harmony and the theoretical optimal solution and the theoretical worst solution respectively.

21. An electronic device, characterized in that: include: at least one processor and memory; The memory and the processor are connected via a bus; The memory is used to store one or more programs; When the one or more programs are executed by the at least one processor, a multi-agent collaborative loss reduction multi-objective optimization method based on a harmony algorithm as described in any one of claims 1 to 10 is implemented.

22. A computing device readable storage medium, characterized in that: An execution program is stored thereon, and when the execution program is executed, a multi-agent collaborative loss reduction multi-objective optimization method based on a harmony algorithm as described in any one of claims 1 to 10 is implemented.

Images