A Multimodal Distributed Multi-Objective Hierarchical Intelligent Integrated Energy System Scheduling Method

Abstract

The present invention proposes a multi-modal distributed multi-objective hierarchical intelligent integrated energy system scheduling method, which combines a multi-layer distributed multi-objective consensus method and a multi-modal multi-objective intelligent method for the intelligent scheduling of the integrated energy system. First, the multi-layer distributed multi-objective consensus method in the proposed method is used to solve the problems of system calculation speed, robustness, and information privacy when the scale of the interconnected integrated energy system increases and the number of agents increases. Second, the multi-modal multi-objective intelligent method in the proposed method is used to solve the multi-modal characteristic problems in multi-modal multi-objective scheduling. The proposed method uses multi-region hierarchical parallel processing to improve the optimization speed of the integrated energy system, solves the problem that one Pareto front in the objective space corresponds to two or more optimal Pareto solution sets in the decision space in the multi-modal multi-objective scheduling problem, and provides diverse alternative solution sets.

Description

Technical Field

[0001] The present invention belongs to the field of dispatching of power systems, new power systems and integrated energy systems, and particularly relates to a multimodal distributed multi-objective hierarchical intelligent dispatching method. Background Art

[0002] Regarding the multimodal distributed multi-objective hierarchical dispatching problem of integrated energy systems, there are multiple local / global optimal Pareto solution sets that meet the constraint conditions in the decision space of multi-objective optimization. However, after the final multi-objective optimization, only one optimal Pareto solution set is obtained, which makes the final solution set lack diversity and cannot meet the needs of decision-makers, resulting in the loss of diversity in the dispatching of large-scale interconnected power systems.

[0003] Therefore, a multimodal distributed multi-objective hierarchical intelligent integrated energy system dispatching method is proposed. This method can retain the characteristics of multimodality. In the multimodal distributed multi-objective hierarchical dispatching problem of integrated energy systems, it can not only accelerate the speed of the system to handle problems, ensure the privacy and robustness of the system, but also provide diverse choices for the dispatching of integrated energy systems. Summary of the Invention

[0004] The present invention proposes a multimodal distributed multi-objective hierarchical intelligent integrated energy system dispatching method, which combines a multi-layer distributed multi-objective consistency method and a multimodal multi-objective intelligent method for the intelligent dispatching of integrated energy systems; the steps of the proposed method in the using process are as follows:

[0005] (1) Construct an optimal power flow model of the integrated energy system, with the power generation cost and carbon emissions as the objectives, and follow the equality constraints of economic dispatching, the power inequality constraints of economic dispatching, and the reserve inequality constraints of economic dispatching;

[0006] The objective functions for minimizing the power generation cost and carbon emissions objectives are:

[0007]

[0008] Wherein, f 1 (x) is the power generation cost; f 2 (x) is the carbon emissions; is the cost of the i-th conventional generator unit at time t; is the cost of the j-th wind turbine generator unit at time t; is the cost of the z-th solar photovoltaic generator unit at time t; is the cost of the k-th hydroelectric generator unit at time t; is the cost of the l-th geothermal generator unit at time t; is the carbon emissions of the i-th conventional generator unit at time t; NGE is the number of conventional generator sets; N WE is the number of wind turbine generator sets; N PE is the number of solar photovoltaic generator sets; N HE is the number of hydro generator sets; N OE is the number of geothermal generator sets; T is the statistical time of the objective function; and it satisfies:

[0009]

[0010] Among them, is the power generation of the i-th conventional generator set at time t; is the power generation of the j-th wind turbine generator set at time t; is the power generation of the z-th solar wind generator set at time t; is the power generation of the k-th hydro generator set at time t; is the power generation of the l-th geothermal generator set at time t; a i is the quadratic term of the cost coefficient of the i-th conventional generator set; b i is the linear term of the cost coefficient of the i-th conventional generator set; c i is the constant term of the cost coefficient of the i-th conventional generator set; d j is the unit power generation economic cost of the j-th wind turbine generator set; e z is the unit power generation economic cost of the z-th solar photovoltaic generator set; g k is the unit power generation economic cost of the k-th hydro generator set; m l is the unit power generation economic cost of the l-th geothermal generator set; α i is the quadratic term of the carbon emission coefficient of the i-th conventional generator set; β i is the linear term of the carbon emission coefficient of the i-th conventional generator set; γ i is the constant term of the carbon emission coefficient of the i-th conventional generator set;

[0011] The equality constraint of economic dispatch is:

[0012]

[0013] The power inequality constraint of economic dispatch is:

[0014]

[0015] Among them, is the predicted load value at time t; is the lower limit of the power generation of the i-th conventional generator set; is the upper limit of the power generation of the i-th conventional generator set; is the lower limit of the power generation of the j-th wind turbine generator unit; is the upper limit of the power generation of the j-th wind turbine generator unit; is the lower limit of the power generation of the z-th solar photovoltaic generator unit; is the upper limit of the power generation of the z-th solar photovoltaic generator unit; is the lower limit of the power generation of the k-th hydraulic generator unit; is the upper limit of the power generation of the k-th hydraulic generator unit; is the lower limit of the power generation of the l-th geothermal generator unit; is the upper limit of the power generation of the l-th geothermal generator unit; is the power generation of the conventional generator unit at the (t - 1)th moment; is the power generation of the conventional generator unit at the t-th moment; is the downward ramp value of the conventional generator unit; is the upward ramp value of the conventional generator unit; T 60 is 60 minutes;

[0016] The reserve inequality constraint for economic dispatch is:

[0017]

[0018] Among them, is the positive spinning reserve value of the wind turbine generator unit at the t-th moment; is the positive spinning reserve value of the j-th wind turbine generator unit at the t-th moment; is the negative spinning reserve value of the wind turbine generator unit at the t-th moment; is the negative spinning reserve value of the j-th wind turbine generator unit at the t-th moment; is the positive spinning reserve value of the solar photovoltaic generator unit at the t-th moment; is the positive spinning reserve value of the z-th solar photovoltaic generator unit at the t-th moment; is the negative spinning reserve value of the solar photovoltaic generator unit at the t-th moment; is the negative spinning reserve value of the z-th solar photovoltaic generator unit at the t-th moment; is the positive spinning reserve value of the hydraulic generator unit at the t-th moment; is the positive spinning reserve value of the k-th hydraulic generator unit at the t-th moment; is the negative spinning reserve value of the hydraulic generator unit at the t-th moment; is the negative spinning reserve value of the k-th hydraulic generator unit at the t-th moment; is the positive spinning reserve value of the geothermal generator unit at the t-th moment; is the positive spinning reserve value of the l-th geothermal generator unit at the t-th moment; is the negative spinning reserve value of the geothermal power generation unit at time t; is the negative spinning reserve value of the l-th geothermal power generation unit at time t; L W+ % is the positive spinning reserve demand coefficient of the load in the wind turbine unit; L W- % is the negative spinning reserve demand coefficient of the load in the wind turbine unit; L P+ % is the positive spinning reserve demand coefficient of the load in the solar photovoltaic power generation unit; L P- % is the negative spinning reserve demand coefficient of the load in the solar photovoltaic power generation unit; L H+ % is the positive spinning reserve demand coefficient of the load in the hydroelectric generator unit; L H- % is the negative spinning reserve demand coefficient of the load in the hydroelectric generator unit; L O+ % is the positive spinning reserve demand coefficient of the load in the geothermal power generation unit; L O- % is the negative spinning reserve demand coefficient of the load in the geothermal power generation unit; W u % is the positive spinning reserve demand coefficient of the wind power generator unit; W d % is the negative spinning reserve demand coefficient of the wind power generator unit; P u % is the positive spinning reserve demand coefficient of the solar photovoltaic power generation unit; P d % is the negative spinning reserve demand coefficient of the solar photovoltaic power generation unit; H u % is the positive spinning reserve demand coefficient of the hydroelectric power generator unit; H d % is the negative spinning reserve demand coefficient of the hydroelectric power generator unit; O u % is the positive spinning reserve demand coefficient of the geothermal power generation unit; O d % is the negative spinning reserve demand coefficient of the geothermal power generation unit; T 10 is 10 minutes;

[0019] (2) Use the multi-modal multi-objective intelligent method to solve the multi-modal characteristics and multi-objective scheduling problems. First, adopt the niche strategy to select more diverse and better individuals as the parental population of each generation; secondly, use a special selection mechanism to increase the selection pressure, improve the diversity, and accelerate the convergence speed; finally, use a special crowding distance to consider the distances of candidate solutions in both the decision space and the objective space, and obtain the optimal Pareto set where the solution sets do not dominate each other, solve the multi-modal characteristics in the scheduling problem, provide more diverse candidate solutions, and at the same time use the characteristics of the knee point and the knee region to provide the most appropriate optimal solution without preference;

[0020] Substitute f 1 (x) and f 2(x) is used as the horizontal axis coordinate and the vertical axis coordinate of the two-dimensional coordinate system respectively. Considering its multi-modal characteristics, after obtaining a series of feasible solutions in the decision space and the target space, multi-modal multi-objective optimization is performed on the feasible solutions to obtain a suitable optimal Pareto solution set. The steps are as follows:

[0021] Step 2.1: Set the initial iteration time t = 1;

[0022] Step 2.2: Take two conflicting objective quantities as the horizontal axis and the vertical axis of the two-dimensional coordinate system and consider them in the same coordinate system;

[0023] Step 2.3: Initialize the parameters, and set the iteration number n of the multi-modal multi-objective to 0;

[0024] Step 2.4: Generate a population of size M within the range of feasible solutions;

[0025] Step 2.5: Embed the niche strategy in the tournament selection to select the mating pool of the parent population;

[0026] Step 2.6: Use the traditional genetic method to perform simulated binary crossover and polynomial mutation operations on the parent population to generate the offspring population, then merge the parent population and the offspring population and perform non-dominated sorting operations;

[0027] Step 2.7: Find the knee point and the knee region on the optimal Pareto front;

[0028] Step 2.8: Perform environmental selection based on non-dominated sorting, identifying the knee point and the knee region, and the special crowding distance to obtain M better individuals;

[0029] Step 2.9: Judge the iteration number. If the iteration number is not satisfied, let n = n + 1 and return to Step 2.5 to continue the iteration until the iteration number is satisfied; if the iteration step is satisfied, then perform Step 2.10;

[0030] Step 2.10: Obtain the optimal Pareto solution set that retains the multi-modal characteristics and select a suitable solution according to the preferences of the decision maker;

[0031] (3) Use the multi-layer distributed multi-objective consistency method to solve the problem of distributed scheduling of the integrated energy system. Divide the integrated energy system into multiple regions, regard each region as an agent for separate optimization. After each region independently completes its own optimization problem, without exchanging the internal variables of each region, only by exchanging the boundary information between regions can the overall optimization be achieved, ensuring the privacy of the system and accelerating the speed of the integrated energy system to solve problems. The steps are as follows:

[0032] Step 3.1: Input the already predicted load data and set the initial iteration step k of each region's optimization to 0;

[0033] Step 3.2: Exchange the boundary variables and target variables in combination with the topological structure of the agent;

[0034] Step 3.3: Calculate the active power output of each unit;

[0035] Step 3.4: Correct the active power output;

[0036] Step 3.5: Calculate the active power deviation value;

[0037] Step 3.6: Judge whether the active power deviation meets the requirements. If it does not meet the requirements, let k = k + 1 and continue to return to Step 3.2 for iteration; if it meets the requirements, end and obtain the optimal power generation of each region;

[0038] (4) Use the multi - layer distributed multi - objective consistency method to solve the hierarchical scheduling problem of the integrated energy system. Regard the regions divided in the first step as the first layer, and re - divide the regions within the regions divided in the first layer. Perform multi - modal multi - objective distributed optimal scheduling on the re - divided regions. When solving problems for the second - layer regions divided within each region in the first layer, they are parallel, which can accelerate the problem - solving speed reduced due to the continuous increase of the integrated energy system; moreover, it can perform multiple hierarchical divisions on the regions, reducing the difficulty of solving problems for the huge and complex integrated energy system. The steps are as follows:

[0039] Step 4.1: Input the optimal power generations calculated for Region 1, Region 2, and Region 3 within the first layer into the second layer as load values respectively;

[0040] Step 4.2: Use the methods from Step 2.1 to Step 2.10 and Step 3.1 to Step 3.6 to process the second - layer regions of Region 1: Region 11, Region 12, Region 13; the second - layer regions of Region 2: Region 21, Region 22, Region 23; and the third - layer regions of Region 3: Region 31, Region 32, Region 33;

[0041] Step 4.3: Input the optimal power generations calculated for Region 11, Region 12, Region 13, Region 21, Region 22, Region 23, Region 31, Region 32, and Region 33 into the third layer as load values respectively;

[0042] Step 4.4: Use the methods in Steps 2.1 to 2.10 and Steps 3.1 to 3.6 to perform parallel processing on the third layer of Region 11: Region 111, Region 112, Region 113; on the third layer of Region 12: Region 121, Region 122, Region 123; on the third layer of Region 13: Region 131, Region 132, Region 133; on the third layer of Region 21: Region 211, Region 212, Region 213; on the third layer of Region 22: Region 221, Region 222, Region 223; on the third layer of Region 23: Region 231, Region 232, Region 233; on the third layer of Region 31: Region 311, Region 312, Region 313; on the third layer of Region 32: Region 321, Region 322, Region 323; on the third layer of Region 33: Region 331, Region 332, Region 333;

[0043] Step 4.5: Take the optimal power generation obtained from the parallel calculation of each region as the load and input it to the next layer, and use the methods in Steps 2.1 to 2.10 and Steps 3.1 to 3.6 to solve it;

[0044] Step 4.6: Determine whether the conditions are met. If the conditions are not met, let t = t + 1 and transfer to Step 2.2 to continue the iteration; if the conditions are met, the iteration ends and the optimal power generation of each region is output.

[0045] The present invention has the following advantages and effects compared with the prior art:

[0046] (1) It can effectively utilize clean energy and solve the serious problems of wind curtailment, light curtailment, and water curtailment in some areas; for the problem that a large amount of clean energy is incorporated into the power grid, resulting in difficult consumption, an effective method is proposed;

[0047] (2) It builds a framework of a multi-modal distributed multi-objective hierarchical intelligent method for the economic dispatching problem of the integrated energy system;

[0048] (3) The concept of multi-modal is introduced. In the previous dispatching of power systems, there are not only multi-objective problems but also multi-modal problems, resulting in multiple solutions that meet the constraint conditions when dealing with problems. However, the multi-modal characteristics are not considered in the previous dispatching;

[0049] (4) The constructed multi-modal distributed multi-objective hierarchical intelligent method solves the multi-modal characteristic problem in the dispatching of the integrated energy system on the basis of improving speed and ensuring privacy. Brief Description of the Drawings

[0050] Figure 1 It is a distributed hierarchical schematic diagram of the method of the present invention.

[0051] Figure 2It is a schematic diagram of distributed multi-modal multi-objectives of the method of the present invention.

[0052] Figure 3 It is a flowchart of the distributed multi-modal multi-objective method of the method of the present invention.

[0053] Figure 4 It is the overall flowchart of the method of the present invention. Detailed implementation manners

[0054] A multi-modal distributed multi-objective hierarchical intelligent integrated energy system scheduling method proposed by the present invention is described in detail with reference to the accompanying drawings as follows:

[0055] Figure 1 It is a distributed hierarchical schematic diagram of the method of the present invention. First, the integrated energy system is divided into three regions, namely Region 1, Region 2, and Region 3, which is the first layer; secondly, Region 1 is divided into Region 11, Region 12, and Region 13, Region 2 is divided into Region 21, Region 22, and Region 23, and Region 3 is divided into Region 31, Region 32, and Region 33, which is the second layer; then the regions in the second layer are further divided into Region 111, Region 112, Region 113, Region 121, Region 122, Region 123, Region 131, Region 132, Region 133, Region 211, Region 212, Region 213, Region 221, Region 222, Region 223, Region 231, Region 232, Region 233, Region 311, Region 312, Region 313, Region 321, Region 322, Region 323, Region 331, Region 332, and Region 333. Each agent in each layer solves problems independently, does not exchange internal information, and only exchanges external information, ensuring the privacy of the system. At the same time, each agent in each layer solves problems in parallel, improving the speed of the integrated energy system to solve problems. The distributed hierarchical scheduling of the integrated energy system is realized.

[0056] Figure 2 It is a schematic diagram of distributed multi-modal multi-objectives of the method of the present invention. Taking the third-layer region inside Region 11 as an example, each region in Region 111, Region 112, and Region 113 is regarded as an agent and solves problems independently. The optimal Pareto front of the objective space of each region corresponds to multiple optimal Pareto solution sets in the decision space, reflecting the multi-modal characteristics. The multi-objective multi-modal intelligent method is used to solve the multi-modal scheduling problem of the integrated energy scheduling system, reflecting the diversity of scheduling.

[0057] Figure 3It is the flowchart of the distributed multi-modal multi-objective method of the method of the present invention. The basic idea of the distributed multi-modal multi-objective method is to retain the multi-modal characteristics of the multi-objective problem, provide the decision maker with a diverse set of optimal Pareto solutions for selection, and perform overall scheduling by exchanging the boundary regions of each agent area. The specific steps are as follows:

[0058] Step 1: Initialize the parameters, and set the number of iterations n of the multi-modal multi-objective optimization to 0;

[0059] Step 2: Consider two conflicting objective quantities as the horizontal and vertical axes of a two-dimensional coordinate system and place them in the same coordinate system;

[0060] Step 3: Generate a population of size M within the range of feasible solutions;

[0061] Step 4: Embed the niche strategy in the tournament selection to select the mating pool of the parent population;

[0062] Step 5: Use the traditional genetic method to perform simulated binary crossover and polynomial mutation operations on the parent population to generate the offspring population, then combine the parent population and the offspring population and perform non-dominated sorting operations;

[0063] Step 6: Find the knee point and knee region on the optimal Pareto front;

[0064] Step 7: Perform environmental selection based on non-dominated sorting, identifying knee points and knee regions, and special crowding distances to obtain M better individuals;

[0065] Step 8: Judge the number of iterations. If the number of iterations is not satisfied, set n = n + 1 and return to Step 4 to continue the iteration until the number of iterations is satisfied; if the number of iterations is satisfied, proceed to the next step;

[0066] Step 9: Obtain the optimal Pareto solution set that retains the multi-modal characteristics and select a suitable solution according to the preferences of the decision maker;

[0067] Step 10: Input the already predicted load data and set the initial iteration step k = 0;

[0068] Step 11: Exchange the boundary variables and objective variables in combination with the topological structure of the agent;

[0069] Step 12: Calculate the active power output of each unit;

[0070] Step 13: Correct the active power output;

[0071] Step 14: Calculate the active power deviation value;

[0072] Step 15: Determine whether the active power deviation meets the requirements. If not, let k = k + 1 and return to Step 11 for iteration; if it meets the requirements, end and obtain the optimal power generation of each region.

[0073] Figure 4 It is the overall flowchart of the method of the present invention. The specific steps are as follows:

[0074] Step 1: Set the initial iteration time t = 1;

[0075] Step 2: Input the predicted load value and set the initial iteration step k = 0;

[0076] Step 3: Divide the integrated energy system into Region 1, Region 2, and Region 3, and regard each region as an agent;

[0077] Step 4: Use the Figure 3 distributed multi-modal multi-objective method in to process Region 1, Region 2, and Region 3 in the first layer;

[0078] Step 5: Input the optimal power generation calculated by Region 1, Region 2, and Region 3 as the load value into the second layer respectively;

[0079] Step 6: Use the Figure 3 distributed multi-modal multi-objective method in to process the second layer of Region 1: Region 11, Region 12, Region 13; the second layer of Region 2: Region 21, Region 22, Region 23; the third layer of Region 3: Region 31, Region 32, Region 33;

[0080] Step 7: Input the optimal power generation calculated by Region 11, Region 12, Region 13, Region 21, Region 22, Region 23, Region 31, Region 32, and Region 33 as the load value into the third layer respectively;

[0081] Step 8: Use the Figure 3 distributed multi-modal multi-objective method in to perform parallel processing on the third layer of Region 11: Region 111, Region 112, Region 113; the third layer of Region 12: Region 121, Region 122, Region 123; the third layer of Region 13: Region 131, Region 132, Region 133; the third layer of Region 21: Region 211, Region 212, Region 213; the third layer of Region 22: Region 221, Region 222, Region 223; the third layer of Region 23: Region 231, Region 232, Region 233; the third layer of Region 31: Region 311, Region 312, Region 313; the third layer of Region 32: Region 321, Region 322, Region 323; the third layer of Region 33: Region 331, Region 332, Region 333;

[0082] Step 9: Use the optimal power generation obtained from the parallel calculations of each region as the load input to the next layer, and solve it using the distributed multi-modal multi-objective method in Figure 3 ;

[0083] Step 10: Determine whether t ≤ T is satisfied. If the condition is not satisfied, let t = t + 1 and go back to Step 2 to continue the iteration; if the condition is satisfied, the iteration ends and the optimal power generation of each region is output.

Claims

1. A multi-modal distributed multi-objective hierarchical intelligent integrated energy system scheduling method, characterized in that, this method combines the multi-layer distributed multi-objective consensus method and the multi-modal multi-objective intelligent method for the intelligent scheduling of the integrated energy system; the steps of the proposed method during use are as follows: (1) Construct an optimal power flow model of the integrated energy system, with the power generation cost and carbon emissions as the objectives, and follow the equality constraints of economic dispatch, the power inequality constraints of economic dispatch, and the reserve inequality constraints of economic dispatch; The objective functions for minimizing the power generation cost and carbon emissions objectives are: Among them, f 1 (x) is the power generation cost; f 2 (x) is the carbon emission; is the cost of the i-th conventional generator unit at time t; is the cost of the j-th wind power generation unit at time t; is the cost of the z-th solar photovoltaic power generation unit at time t; is the cost of the k-th hydraulic power generation unit at time t; is the cost of the l-th geothermal power generation unit at time t; is the carbon emission of the i-th conventional generator unit at time t; N GE is the number of conventional generator units; N WE is the number of wind power generation units; N PE is the number of solar photovoltaic power generation units; N HE is the number of hydraulic power generation units; N OE is the number of geothermal power generation units; T is the statistical time of the objective function; and it satisfies: Among them, is the power generation of the i-th conventional generator unit at time t; is the power generation of the j-th wind turbine generator unit at time t; is the power generation of the z-th solar and wind power generation unit at time t; is the power generation of the k-th hydroelectric generator unit at time t; is the power generation of the l-th geothermal power generation unit at time t; a i is the quadratic term of the cost coefficient of the i-th conventional generator unit; b i is the linear term of the cost coefficient of the i-th conventional generator unit; c i is the constant term of the cost coefficient of the i-th conventional generator unit; d j is the unit power generation economic cost of the j-th wind turbine generator unit; e z is the unit power generation economic cost of the z-th solar photovoltaic power generation unit; g k is the unit power generation economic cost of the k-th hydroelectric generator unit; m l is the unit power generation economic cost of the l-th geothermal power generation unit; α i is the quadratic term of the carbon emission coefficient of the i-th conventional generator unit; β i is the linear term of the carbon emission coefficient of the i-th conventional generator unit; γ i is the constant term of the carbon emission coefficient of the i-th conventional generator unit; The equality constraints of economic dispatch are: The power inequality constraints of economic dispatch are: Among them, is the predicted load value at time t; is the lower limit of the power generation of the i-th conventional generator unit; is the upper limit of the power generation of the i-th conventional generator unit; is the lower limit of the power generation of the j-th wind power generation unit; is the upper limit of the power generation of the j-th wind power generation unit; is the lower limit of the power generation of the z-th solar photovoltaic power generation unit; is the upper limit of the power generation of the z-th solar photovoltaic power generation unit; is the lower limit of the power generation of the k-th hydroelectric generator unit; is the upper limit of the power generation of the k-th hydroelectric generator unit; is the lower limit of the power generation of the l-th geothermal power generation unit; is the upper limit of the power generation of the l-th geothermal power generation unit; is the power generation of the conventional generator unit at time t-1; is the power generation of the conventional generator unit at time t; is the downward ramp value of the conventional generator unit; is the upward ramp value of the conventional generator unit; T 60 is 60 minutes; The reserve inequality constraints of economic dispatch are: Among them, is the positive spinning reserve value of the wind turbine generator set at time t; is the positive spinning reserve value of the j-th wind turbine generator set at time t; is the negative spinning reserve value of the wind turbine generator set at time t; is the negative spinning reserve value of the j-th wind turbine generator set at time t; is the positive spinning reserve value of the solar photovoltaic generator set at time t; is the positive spinning reserve value of the z-th solar photovoltaic generator set at time t; is the negative spinning reserve value of the solar photovoltaic generator set at time t; is the negative spinning reserve value of the z-th solar photovoltaic generator set at time t; is the positive spinning reserve value of the hydroelectric generator set at time t; is the positive spinning reserve value of the k-th hydroelectric generator set at time t; is the negative spinning reserve value of the hydroelectric generator set at time t; is the negative spinning reserve value of the k-th hydroelectric generator set at time t; is the positive spinning reserve value of the geothermal generator set at time t; is the positive spinning reserve value of the l-th geothermal generator set at time t; is the negative spinning reserve value of the geothermal generator set at time t; is the negative spinning reserve value of the l-th geothermal generator set at time t; L W+ % is the positive spinning reserve demand coefficient of the load in the wind turbine generator set; L W- % is the negative spinning reserve demand coefficient of the load in the wind turbine generator set; L P+ % is the positive spinning reserve demand coefficient of the load in the solar photovoltaic generator set; L P- % is the negative spinning reserve demand coefficient of the load in the solar photovoltaic generator set; L H+ % is the positive spinning reserve demand coefficient of the load in the hydroelectric generator set; L H- % is the negative spinning reserve demand coefficient of the load in the hydroelectric generator set; L O+ % is the positive spinning reserve demand coefficient of the load in the geothermal generator set; L O- % is the negative spinning reserve demand coefficient of the load in the geothermal generator set; W u % is the positive spinning reserve demand coefficient of the wind power generator set; W d % is the negative spinning reserve demand coefficient of the wind power generator set; P u % is the positive spinning reserve demand coefficient of the solar photovoltaic generator set; P d % is the negative spinning reserve demand coefficient of the solar photovoltaic generator set; H u % is the positive spinning reserve demand coefficient of the hydro-generator unit; H d % is the negative spinning reserve demand coefficient of the hydro-generator unit; O u % is the positive spinning reserve demand coefficient of the geothermal generator unit; O d % is the negative spinning reserve demand coefficient of the geothermal generator unit; T 10 is 10 minutes; (2) Step 2.1: Set the initial iteration time t = 1; Step 2.2: Consider the two objective quantities as the horizontal and vertical axes of a two-dimensional coordinate system and place them in the same coordinate system for consideration; Step 2.3: Initialize the parameters, and set the iteration number n of the multi-modal multi-objective to 0; Step 2.4: Generate a population of size M within the range of feasible solutions; Step 2.5: Embed the niche strategy in the tournament selection to select the mating pool of the parent population; Step 2.6: Use the genetic method to perform simulated binary crossover and polynomial mutation operations on the parent population to generate the offspring population, and then merge the parent population and the offspring population for non-dominated sorting operations; Step 2.7: Find the knee point and knee region on the optimal Pareto front; Step 2.8: Perform environmental selection based on non-dominated sorting, identifying the knee point and knee region, and the special crowding distance to obtain M better individuals; Step 2.9: Judge the iteration number. If the iteration number is not satisfied, let n = n + 1 and return to Step 2.5 to continue the iteration until the iteration number is satisfied; if the iteration step is satisfied, then perform Step 2.10; Step 2.10: Obtain the optimal Pareto solution set that retains the multi-modal characteristics; (3) Step 3.1: Input the load data and set the initial iteration step k = 0 for each region's optimization; Step 3.2: Combine the topological structure of the agents to exchange the boundary variables and objective variables; Step 3.3: Calculate the active power output of each unit; Step 3.4: Correct the active power output; Step 3.5: Calculate the active power deviation value; Step 3.6: Judge whether the active power deviation meets the requirements. If it does not meet the requirements, let k = k + 1 and continue to return to Step 3.2 for iteration; if it meets the requirements, then end and obtain the best power generation of each region; (4) Use the multi-layer distributed multi-objective consensus method to solve the problem of hierarchical scheduling of the integrated energy system. Regard the regions divided in the first step as the first layer, and perform further regional division within the regions divided in the first layer. Perform multi-modal multi-objective distributed optimization scheduling on the regions obtained from the further division. The second-layer regions divided within each region of the first layer solve the problem in parallel, which can accelerate the problem-solving speed reduced due to the continuous increase of the integrated energy system; and it is possible to perform multiple hierarchical divisions on the regions, reducing the difficulty of solving problems in the integrated energy system. The steps are as follows: Step 4.1: Input the optimal power generation amounts calculated for Region 1, Region 2, and Region 3 within the first layer into the second layer respectively as load values; Step 4.2: Process Region 11, Region 12, and Region 13 in the second layer of Region 1; Region 21, Region 22, and Region 23 in the second layer of Region 2; and Region 31, Region 32, and Region 33 in the third layer of Region 3 using the methods in Steps 2.1 to 2.10 and Steps 3.1 to 3.6; Step 4.3: Input the optimal power generation amounts calculated for Region 11, Region 12, Region 13, Region 21, Region 22, Region 23, Region 31, Region 32, and Region 33 into the third layer respectively as load values; Step 4.4: Parallelly process Region 111, Region 112, and Region 113 in the third layer of Region 11; Region 121, Region 122, and Region 123 in the third layer of Region 12; Region 131, Region 132, and Region 133 in the third layer of Region 13; Region 211, Region 212, and Region 213 in the third layer of Region 21; Region 221, Region 222, and Region 223 in the third layer of Region 22; Region 231, Region 232, and Region 233 in the third layer of Region 23; Region 311, Region 312, and Region 313 in the third layer of Region 31; Region 321, Region 322, and Region 323 in Region 32; and Region 331, Region 332, and Region 333 in the third layer of Region 33 using the methods in Steps 2.1 to 2.10 and Steps 3.1 to 3.6; Step 4.5: Input the optimal power generation amounts obtained from parallel calculations for each region into the next layer as load, and solve using the methods in Steps 2.1 to 2.10 and Steps 3.1 to 3.6; Step 4.6: Determine whether the condition is met. If the condition is not met, let t = t + 1 and transfer to Step 2.2 to continue the iteration; if the condition is met, end the iteration and output the optimal power generation amounts for each region.

Images