A method and device for collaborative planning of deep peak regulation and composite energy storage for thermal power units

Abstract

The present invention provides a method for collaborative planning of deep peak shaving and composite energy storage for thermal power units, comprising: establishing a multi-objective optimization model for collaborative planning of deep peak shaving and composite energy storage for thermal power units based on optimal economic efficiency, optimal carbon emission reduction, and minimum wind and solar power curtailment; preprocessing input data based on the entropy weight-ideal method, calculating the comprehensive score of each decision variable, and determining the order of inputting deep peak shaving, pumped storage, and chemical energy storage; and using an improved NSGA-II method combined with fuzzy membership to obtain the deep peak shaving amount of thermal power units and the planned capacity of newly built composite energy storage. The present invention proposes an improved NSGA-II algorithm for data preprocessing based on the entropy weight-ideal solution, establishing a multi-objective planning model to obtain the optimal configuration plan for flexible resources. The model has fast solution speed and strong adaptability, takes into account both carbon emission reduction and new energy absorption capacity, significantly reduces the system's wind and solar power curtailment rate, and is conducive to the low-carbon economic operation of the power system.

Description

Technical Field

[0001] The present invention relates to the technical field of power system planning, and specifically but not limited to, a method and device for collaborative planning of deep peak regulation and composite energy storage for thermal power units. Background Art

[0002] Due to the volatility, intermittency, and uncertainty of renewable energy generation, the large-scale integration of renewable energy into the grid has reduced the grid's power supply availability, placing significant pressure on the power system's peak load regulation. In the future, three types of peak load regulation resources—flexible thermal power conversion, chemical energy storage, and pumped hydro—will each account for a significant portion of the grid. Coordinating these three resources will facilitate power balance and ensure safe and stable operation.

[0003] Currently, flexibility resource optimization planning often targets economic optimization or maximizing renewable energy consumption. However, deep peak shaving of thermal power units can lead to reduced operating efficiency and incomplete boiler combustion, creating new environmental pressures. Therefore, it is necessary to balance the flexibility improvements of thermal power units with their ability to reduce carbon emissions.

[0004] In view of this, it is necessary to provide a new structure or method to solve at least part of the above problems. Summary of the Invention

[0005] In response to one or more problems in the prior art, the present invention proposes a method for collaborative planning of deep peak regulation and composite energy storage for thermal power units, proposes an improved NSGA-II algorithm based on entropy weight-ideal solution for data preprocessing, and establishes a multi-objective planning model to obtain the optimal configuration plan for flexibility resources. The method has a fast solution speed and strong adaptability, takes into account both carbon emission reduction and new energy absorption capacity, significantly reduces the system's wind and solar power curtailment rate, and is conducive to the low-carbon economic operation of the power system.

[0006] The technical solutions for achieving the purpose of the present invention are:

[0007] A method for collaborative planning of deep peak regulation and composite energy storage for thermal power units, comprising:

[0008] S1. Establish an economic model with the goal of maximizing annual revenue, a carbon emission reduction model with the goal of maximizing carbon emission reduction, and a wind and solar curtailment model with the goal of minimizing wind and solar curtailment. Based on the aforementioned economic model, carbon emission reduction model, and wind and solar curtailment model, establish a multi-objective optimization model for deep peak regulation of thermal power units and coordinated planning of composite energy storage.

[0009] S2. Preprocessing input data based on the entropy weight-ideal method, where the input data is each decision variable, including thermal power deep peak regulation, pumped storage, and chemical energy storage, calculating a comprehensive score for each decision variable in the multi-objective optimization model, and determining the order of inputting thermal power deep peak regulation, pumped storage, and chemical energy storage according to the comprehensive score of each decision variable from largest to smallest;

[0010] S3. An improved NSGA-II method is used to solve the model. The improved method generates offspring by introducing a normal distribution crossover operator and an adaptive adjustment mutation method based on the NSGA-II method, obtains the Pareto optimal solution set, and obtains the comprehensive optimal solution through fuzzy membership, thereby obtaining the deep peak regulation amount of the thermal power unit and the planned capacity of the newly built composite energy storage.

[0011] Furthermore, in the method for collaborative planning of deep peak regulation and composite energy storage for thermal power units of the present invention, the economic model established in S1 with the goal of maximizing annual revenue is:

[0012] max F1=Pr t +Pr n -C H -C s

[0013] Where, F1 is the annual income; Pr t and Pr n are thermal power generation income and new energy power generation income respectively; C H with C s are the cost of pumped storage units and chemical energy storage respectively;

[0014] Among them, thermal power revenue Pr t This includes the power grid company's revenue from thermal power sales and the grid company's compensation to units participating in deep peak regulation, specifically:

[0015]

[0016] Where n is the total number of sampling points per year; i represents the i-th moment; P t is the actual output k of the thermal power unit s and k b are the grid electricity sales price and the thermal power on-grid benchmark price respectively; P td k is the power generated by the thermal power unit less than the lower limit of normal output, that is, the power participating in deep peak regulation; td is the compensation coefficient for deep peak-shaving electricity price of thermal power;

[0017] New energy power generation income Pr n The revenue from electricity sales, after deducting the cost of abandoned wind and solar power, is as follows:

[0018]

[0019] Where, P n is the total output of grid-connected new energy; P d is the abandoned wind and solar power; k d is the cost coefficient of abandoned wind and solar power;

[0020] Pumped storage unit cost C H This includes the construction cost of new pumped storage units and the annual operation and maintenance costs of all pumped storage units, specifically:

[0021] C H =P H k H +P nH k nH / m H

[0022] Where, P H and k H are the total installed capacity of the pumped storage units and their annual operation and maintenance cost coefficients; P nH and k nH are the newly added pumped storage power and the construction cost coefficient of the pumped storage unit; m H The service life of the pumped storage unit;

[0023] Chemical energy storage cost C s Including new construction costs and operation and maintenance costs, specifically:

[0024]

[0025] Where, E es is the newly added chemical energy storage capacity; k es is the construction cost coefficient of chemical energy storage; m s and n s are the annual charge and discharge times of chemical energy storage and the total charge and discharge times in the entire life cycle; ν is the annual operation and maintenance cost coefficient of chemical energy storage.

[0026] Furthermore, in the method for collaborative planning of deep peak regulation and composite energy storage for thermal power units of the present invention, a carbon emission reduction model is established in S1 with the goal of maximizing carbon emission reduction as the following:

[0027] max F2=T p +T s -T t

[0028]

[0029] Where F2 is the carbon emission reduction, which is the carbon emission reduction brought by the composite energy storage minus the carbon emission increase brought by the deep peak regulation of the coal-fired power units; T p and T s are the carbon emission reductions converted from the operation of pumped storage units and chemical energy storage; Tt The incremental carbon emissions from deep peak regulation of coal-fired power units; k nf P is the carbon emission coefficient per unit power of the thermal power unit in normal operation; nf k is the output of thermal power units that have not entered deep peak regulation; df P is the unit carbon emission increment coefficient under deep peak regulation of thermal power units; p and P s are the real-time power of the pumped storage unit and the real-time power of chemical energy storage respectively; P t is the actual output of the thermal power unit; μ gen and μ n d They are all 0-1 variables, and a value of 1 indicates that the pumped storage unit is generating electricity or the chemical energy storage is discharging.

[0030] Furthermore, in the method for collaborative planning of deep peak regulation and composite energy storage for thermal power units of the present invention, in S1, a wind and solar power abandonment model is established with the goal of minimizing the amount of wind and solar power abandoned as follows:

[0031]

[0032] Where F3 is the amount of wind and solar power curtailment, P d To abandon wind and solar power;

[0033] Among them, the constraints of the wind and solar curtailment model include:

[0034] 1) System power balance constraints:

[0035]

[0036] Where, i represents the i-th moment; P t is the actual output of the thermal power unit; P n is the total output of grid-connected new energy; μ pum and μ gen are all 0-1 variables and satisfy μ pum μ gen =0, μ pum is the pumping sign, μ pum =1 means pumping water, μ gen is the power generation symbol, μ gen =1 indicates power generation; and are all 0-1 variables and satisfy μ n c A symbol for charging chemical energy storage. When it is charging, It is the discharge mark of chemical energy storage. When it is discharged; P p and P sare the real-time power of the pumped storage unit and the real-time power of chemical energy storage respectively; P d P is the abandoned wind and solar power; L is the system load; P tie is the inter-provincial exchange power, with outflow being positive;

[0037] 2) Battery energy storage power balance constraints:

[0038]

[0039] E s,min ≤E s (i)≤E s,max

[0040] Where, E s (i) is the amount of electricity stored in the battery at time i; T s is the unit time; η c and η d are the charging and discharging efficiencies of battery energy storage; E s,min With E s,max They are the lower limit and upper limit of battery energy storage capacity respectively;

[0041] 3) Battery energy storage output constraints:

[0042]

[0043] Where, P sn The rated power of the battery energy storage;

[0044] 4) Pumped storage unit capacity constraints:

[0045]

[0046] E H,min ≤E H (i)≤E H,max

[0047] Where, E H (i) is the water storage capacity of the pumped storage power station at time i; E H,min and E H,max are the minimum and maximum water storage capacities of the pumped storage power station respectively; η pum and η gen are the pumping and power generation efficiencies, respectively;

[0048] 5) Output constraints of pumped storage units:

[0049]

[0050] Furthermore, the method for collaborative planning of deep peak regulation and composite energy storage for thermal power units of the present invention, specifically step S2 includes:

[0051] S2-1. Normalize the economic target, carbon emission reduction target, or wind and solar curtailment target function value of each decision variable as follows:

[0052]

[0053] Where x ij is the jth objective function value of the i-th decision variable, p ij is x ij The normalized result of ; m and n represent the number of objective functions and the number of decision variables respectively;

[0054] S2-2. Calculate the information entropy e of the objective function value of each decision variable j , utility value d j and entropy weight w j , specifically:

[0055]

[0056] S2-3. Use the ideal solution method to determine the optimal solution. For positive ideal solutions such as economic targets and carbon emission reduction targets, there are:

[0057]

[0058] For the negative ideal solution of the wind power curtailment target, we have:

[0059]

[0060] Where x j It refers to the set of the jth objective function values ​​of all decision variables, that is, x j ={x 1j ,x 2j ,x 3j}; The positive ideal solution is used to process x ij The larger the better the indicator, the negative ideal solution is used to process x ij The smaller the better the index, the better it is, and the larger the better the index x′ ij ;

[0061] S2-4. Establish a normalized matrix:

[0062]

[0063] z ij =x′ ij ×w j

[0064] Where x′ ij Refers to all ideal solutions, including positive ideal solutions and negative ideal solutions, Z is the comprehensive weighted index zij The matrix formed;

[0065] S2-5. Calculate the optimal solution, the worst solution, the optimal distance, and the worst distance of the i-th decision variable, specifically:

[0066]

[0067] Where z + is the optimal solution, z - is the worst solution, is the optimal distance, is the worst distance;

[0068] S2-6. Calculate the weighted comprehensive score of the i-th variable, specifically:

[0069]

[0070] Where C i represents the comprehensive score of the i-th decision variable, with a value range of (0,1), C i The closer it is to 1, the closer the indicators of the decision variables are to the optimal level, and the higher the comprehensive score is.

[0071] Furthermore, in the collaborative planning method of deep peak regulation and composite energy storage for thermal power units of the present invention, the improved NSGA-II method in S3 is:

[0072] (1) The initialization algorithm generates a parent population of n individuals, i.e., the newly created amount of each peak-shaving means, and calculates the objective function value according to the order of input of each peak-shaving means determined by data preprocessing;

[0073] (2) Introducing the normal distribution crossover operator and the adaptive adjustment mutation method to perform crossover mutation to produce an offspring with n individuals;

[0074] (3) Merge the parent and child populations, and perform non-dominated sorting and crowding calculation on the objective function value of the merged population;

[0075] (4) Select the first n individuals to generate a new parent population;

[0076] (5) Check whether the stopping condition is met, that is, all individuals are non-dominated solutions. If not, return to step (2). If so, output the Pareto optimal solution set.

[0077] Furthermore, in the method for collaborative planning of deep peak regulation and composite energy storage for thermal power units of the present invention, obtaining a comprehensive optimal solution through fuzzy membership in S3 includes:

[0078] The fuzzy membership function when solving the objective function maximization problem is:

[0079]

[0080] The fuzzy membership function when solving the objective function minimization problem is:

[0081]

[0082] Where, f j represents the j-th objective function value; and Indicates the maximum and minimum values ​​of the jth objective function; objective functions F1 and F2 are respectively the maximum total system benefit and the maximum carbon emission reduction, and the maximum membership function is selected; objective function F3 is the minimum amount of abandoned wind and solar power, and the minimum membership function is selected;

[0083] The satisfaction of all individuals in the Pareto optimal solution set is:

[0084]

[0085] In the formula, h represents overall satisfaction, and a larger h value indicates a higher satisfaction.

[0086] A device for collaborative planning of deep peak regulation and composite energy storage for thermal power units, comprising:

[0087] The data acquisition unit is used to obtain the output data, load and external power data of wind power, photovoltaic power and synchronous generators in a certain area at the same time for one year or multiple years, and to obtain the planning data of new energy, pumped storage units and chemical energy storage in future years;

[0088] A modeling unit is used to establish a multi-objective optimization model for deep peak regulation of thermal power units and coordinated planning of composite energy storage based on economic efficiency, carbon emission reduction, and curtailed wind and solar power, and to set corresponding constraints;

[0089] The solving unit is used to solve the multi-objective optimization model using the improved NSGA-II algorithm that pre-processes data based on the entropy weight ideal solution, obtain the Pareto optimal solution set, and obtain the comprehensive optimal solution through fuzzy membership, thereby obtaining the deep peak regulation amount of the thermal power unit and the planned capacity of the newly built composite energy storage.

[0090] Compared with the prior art, the present invention adopts the above technical solution and has the following technical effects:

[0091] 1. The collaborative planning method for deep peak regulation and composite energy storage of thermal power units of the present invention takes into account both carbon emission reduction and new energy absorption capacity, which can significantly reduce the system's wind and solar power curtailment rate, and is conducive to the low-carbon economic operation of the power system.

[0092] 2. The improved NSGA-II algorithm proposed in the collaborative planning method of deep peak regulation and composite energy storage of thermal power units of the present invention for preprocessing data based on the entropy weight-ideal solution has a faster solution speed and strong adaptability. BRIEF DESCRIPTION OF THE DRAWINGS

[0093] The accompanying drawings are used to provide a further understanding of the present invention and, together with the description, to explain the embodiments of the present invention, but do not constitute a limitation of the present invention. In the accompanying drawings:

[0094] Figure 1 A flow chart of the method for collaborative planning of deep peak regulation and composite energy storage for thermal power units of the present invention is shown.

[0095] Figure 2 A comparison diagram of the iterative processes of the improved NSGA-II algorithm and the traditional NSGA-II algorithm for the collaborative planning method of deep peak regulation and composite energy storage of thermal power units of the present invention is shown.

[0096] Figure 3 The Pareto optimal solution set and comprehensive optimal solution of the multi-objective programming of the improved NSGA-Ⅱ algorithm of the deep peak regulation and composite energy storage collaborative planning method of the thermal power unit of the present invention are shown. DETAILED DESCRIPTION

[0097] In order to further understand the present invention, preferred embodiments of the present invention are described below in conjunction with examples. However, it should be understood that these descriptions are only for further illustrating the features and advantages of the present invention, rather than limiting the claims of the present invention.

[0098] The description in this section is based on typical embodiments only, and the present invention is not limited to the scope of the embodiments described. Combinations of different embodiments, replacement of certain technical features in different embodiments, and replacement of certain technical features in the embodiments with the same or similar prior art methods are also within the scope of the present invention.

[0099] According to one aspect of the present invention, a method for collaborative planning of deep peak regulation and composite energy storage for thermal power units includes:

[0100] S1. Establish an economic model with the goal of maximizing annual revenue, a carbon emission reduction model with the goal of maximizing carbon emission reduction, and a wind and solar curtailment model with the goal of minimizing wind and solar curtailment. Based on the aforementioned economic model, carbon emission reduction model, and wind and solar curtailment model, establish a multi-objective optimization model for deep peak regulation of thermal power units and coordinated planning of composite energy storage. Among them:

[0101] (1) The economic model is established with the goal of maximizing annual revenue as follows:

[0102] max F1=Pr t +Pr n -CH -C s

[0103] Where, F1 is the annual income; Pr t and Pr n are thermal power generation income and new energy power generation income respectively; C H with C s are the cost of pumped storage units and chemical energy storage respectively;

[0104] Among them, thermal power revenue Pr t This includes the power grid company's revenue from thermal power sales and the grid company's compensation to units participating in deep peak regulation, specifically:

[0105]

[0106] Where n is the total number of sampling points per year; i represents the i-th moment; P t is the actual output k of the thermal power unit s and k b are the grid electricity sales price and the thermal power on-grid benchmark price respectively; P td k is the power generated by the thermal power unit less than the lower limit of normal output, that is, the power participating in deep peak regulation; td is the compensation coefficient for deep peak-shaving electricity price of thermal power;

[0107] New energy power generation income Pr n The revenue from electricity sales, after deducting the cost of abandoned wind and solar power, is as follows:

[0108]

[0109] Where, P n is the total output of grid-connected new energy; P d is the abandoned wind and solar power; k d is the cost coefficient of abandoned wind and solar power;

[0110] Pumped storage unit cost C H This includes the construction cost of new pumped storage units and the annual operation and maintenance costs of all pumped storage units, specifically:

[0111] C H =P H k H +P nH k nH / m H

[0112] Where, P H and k H are the total installed capacity of the pumped storage units and their annual operation and maintenance cost coefficients; P nH and k nH are the newly added pumped storage power and the construction cost coefficient of the pumped storage unit; mH The service life of the pumped storage unit;

[0113] Chemical energy storage cost C s Including new construction costs and operation and maintenance costs, specifically:

[0114]

[0115] Where, E es is the newly added chemical energy storage capacity; k es is the construction cost coefficient of chemical energy storage; m s and n s are the annual charge and discharge times of chemical energy storage and the total charge and discharge times in the entire life cycle; ν is the annual operation and maintenance cost coefficient of chemical energy storage.

[0116] (2) The carbon emission reduction model is established with the goal of maximizing carbon emission reduction:

[0117] max F2=T p +T s -T t

[0118]

[0119] Where F2 is the carbon emission reduction, which is the carbon emission reduction brought by the composite energy storage minus the carbon emission increase brought by the deep peak regulation of the coal-fired power units; T p and T s are the carbon emission reductions converted from the operation of pumped storage units and chemical energy storage; T t The incremental carbon emissions from deep peak regulation of coal-fired power units; k nf P is the carbon emission coefficient per unit power of the thermal power unit in normal operation; nf k is the output of thermal power units that have not entered deep peak regulation; df P is the unit carbon emission increment coefficient under deep peak regulation of thermal power units; p and P s are the real-time power of the pumped storage unit and the real-time power of chemical energy storage respectively; P t is the actual output of the thermal power unit; μ gen and They are all 0-1 variables, and a value of 1 indicates that the pumped storage unit is generating electricity or the chemical energy storage is discharging.

[0120] (3) The wind power curtailment model is established with the goal of minimizing the amount of wind power curtailment:

[0121]

[0122] Where F3 is the amount of wind and solar power curtailment, P d To abandon wind and solar power;

[0123] Among them, the constraints of the wind and solar curtailment model include:

[0124] 1) System power balance constraints:

[0125]

[0126] Where, i represents the i-th moment; P t is the actual output of the thermal power unit; P n is the total output of grid-connected new energy; μ pum and μ gen are all 0-1 variables and satisfy μ pum μ gen =0, μ pum is the pumping sign, μ pum =1 means pumping water, μ gen is the power generation symbol, μ gen =1 indicates power generation; and μ n d are all 0-1 variables and satisfy A symbol for charging chemical energy storage. When it is charging, It is the discharge mark of chemical energy storage. When it is discharged; P p and P s are the real-time power of the pumped storage unit and the real-time power of chemical energy storage respectively; P d P is the abandoned wind and solar power; L is the system load; P tie is the inter-provincial exchange power, with outflow being positive;

[0127] 2) Battery energy storage power balance constraints:

[0128]

[0129] E s,min ≤E s (i)≤E s,max

[0130] Where, E s (i) is the amount of electricity stored in the battery at time i; T s is the unit time; η c and η d are the charging and discharging efficiencies of battery energy storage; E s,min With E s,max They are the lower limit and upper limit of battery energy storage capacity respectively;

[0131] 3) Battery energy storage output constraints:

[0132]

[0133] Where, P sn The rated power of the battery energy storage;

[0134] 4) Pumped storage unit capacity constraints:

[0135]

[0136] E H,min ≤E H (i)≤E H,max

[0137] Where, E H (i) is the water storage capacity of the pumped storage power station at time i; E H,min and E H,max are the minimum and maximum water storage capacities of the pumped storage power station respectively; η pum and η gen are the pumping and power generation efficiencies, respectively;

[0138] 5) Output constraints of pumped storage units:

[0139]

[0140] S2. Preprocess the input data based on the entropy weight-ideal method. The input data is various decision variables, including thermal power deep peak regulation, pumped storage, and chemical energy storage. Calculate the comprehensive score of each decision variable in the multi-objective optimization model, and determine the order of implementation of thermal power deep peak regulation, pumped storage, and chemical energy storage based on the comprehensive score of each decision variable from large to small. The specific steps include:

[0141] S2-1. Normalize the economic target, carbon emission reduction target, or wind and solar curtailment target function value of each decision variable as follows:

[0142]

[0143] Where x ij is the jth objective function value of the i-th decision variable, p ij is x ij The normalized result of ; m and n represent the number of objective functions and the number of decision variables respectively;

[0144] S2-2. Calculate the information entropy e of the objective function value of each decision variable j , utility value d j and entropy weight w j , specifically:

[0145]

[0146] S2-3. Use the ideal solution method to determine the optimal solution. For positive ideal solutions such as economic targets and carbon emission reduction targets, there are:

[0147]

[0148] For the negative ideal solution of the wind power curtailment target, we have:

[0149]

[0150] Where x j It refers to the set of the jth objective function values ​​of all decision variables, that is, x j ={x 1j ,x 2j ,x 3j}; The positive ideal solution is used to process x ij The larger the better the indicator, the negative ideal solution is used to process x ij The smaller the better the index, the better it is, and the larger the better the index x′ ij ;

[0151] S2-4. Establish a normalized matrix:

[0152]

[0153] z ij =x′ ij ×w j

[0154] Where x′ ij Refers to all ideal solutions, including positive ideal solutions and negative ideal solutions, Z is the comprehensive weighted index z ij The matrix formed;

[0155] S2-5. Calculate the optimal solution, the worst solution, the optimal distance, and the worst distance of the i-th decision variable, specifically:

[0156]

[0157] Where z + is the optimal solution, z - is the worst solution, is the optimal distance, is the worst distance;

[0158] S2-6. Calculate the weighted comprehensive score of the i-th variable, specifically:

[0159]

[0160] Where C i represents the comprehensive score of the i-th decision variable, with a value range of (0,1), C iThe closer it is to 1, the closer the indicators of the decision variables are to the optimal level, and the higher the comprehensive score is.

[0161] S3. The improved NSGA-II method is used to solve the model. The improved method generates offspring by introducing a normal distribution crossover operator and an adaptive adjustment mutation method based on the NSGA-II method, obtains the Pareto optimal solution set, and obtains the comprehensive optimal solution through fuzzy membership, obtains the deep peak regulation capacity of the thermal power unit and the planned capacity of the new composite energy storage. The improved NSGA-II method is specifically as follows:

[0162] (1) The initialization algorithm generates a parent population of n individuals, i.e., the newly created amount of each peak-shaving means, and calculates the objective function value according to the order of input of each peak-shaving means determined by data preprocessing;

[0163] (2) Introducing the normal distribution crossover operator and the adaptive adjustment mutation method to perform crossover mutation to produce an offspring with n individuals;

[0164] (3) Merge the parent and child populations, and perform non-dominated sorting and crowding calculation on the objective function value of the merged population;

[0165] (4) Select the first n individuals to generate a new parent population;

[0166] (5) Check whether the stopping condition is met, that is, all individuals are non-dominated solutions. If not, return to step (2). If so, output the Pareto optimal solution set.

[0167] Among them, the comprehensive optimal solutions obtained through fuzzy membership include:

[0168] The fuzzy membership function when solving the objective function maximization problem is:

[0169]

[0170] The fuzzy membership function when solving the objective function minimization problem is:

[0171]

[0172] Where, f j represents the j-th objective function value; and Indicates the maximum and minimum values ​​of the jth objective function; objective functions F1 and F2 are respectively the maximum total system benefit and the maximum carbon emission reduction, and the maximum membership function is selected; objective function F3 is the minimum amount of abandoned wind and solar power, and the minimum membership function is selected;

[0173] The satisfaction of all individuals in the Pareto optimal solution set is:

[0174]

[0175] In the formula, h represents overall satisfaction, and a larger h value indicates a higher satisfaction.

[0176] According to another aspect of the present invention, a device for collaborative planning of deep peak regulation and composite energy storage for thermal power units comprises:

[0177] The data acquisition unit is used to obtain the output data, load and external power data of wind power, photovoltaic power and synchronous generators in a certain area at the same time for one year or multiple years, and to obtain the planning data of new energy, pumped storage units and chemical energy storage in future years;

[0178] A modeling unit is used to establish a multi-objective optimization model for deep peak regulation of thermal power units and coordinated planning of composite energy storage based on economic efficiency, carbon emission reduction, and curtailed wind and solar power, and to set corresponding constraints;

[0179] The solving unit is used to solve the multi-objective optimization model using the improved NSGA-II algorithm that pre-processes data based on the entropy weight ideal solution, obtain the Pareto optimal solution set, and obtain the comprehensive optimal solution through fuzzy membership, thereby obtaining the deep peak regulation amount of the thermal power unit and the planned capacity of the newly built composite energy storage.

[0180] Example 1

[0181] The present invention preprocesses the data before calculating the objective function value using the improved NSGA-II algorithm, proposes several common evaluation indicators for all decision variables (the algorithm objective function can also be used as an evaluation indicator), calculates the comprehensive evaluation score of each decision variable based on the entropy weight-ideal method, determines the priority of each decision variable from high to low, simplifies the calculation of the objective function according to the priority, and thus reduces the number of algorithm iterations and the amount of calculation.

[0182] This example uses data from a provincial power grid for 2025 to coordinate deep peak-shaving of thermal power units and integrated energy storage. By 2022, 3,500 MW of pumped-storage capacity had been installed, and by 2025, the province could have built a maximum of 2,500 MW of new pumped-storage capacity. Grid-connected renewable energy capacity reached approximately 38,000 MW, including 9,000 MW of wind power and 29,000 MW of photovoltaic power, representing approximately 40% of installed capacity. Based on the grid's load and actual renewable energy output data from 2020-2021, the grid's load and renewable energy output data for 2025 were projected.

[0183] As of 2022, the provincial grid's installed thermal power capacity will be approximately 57,000 MW. Due to the increasing integration of renewable energy, thermal power units are gradually being shut down, and some units are aging and cannot be retrofitted for flexibility. Therefore, it is estimated that up to 20,000 MW of thermal power units will be eligible for varying degrees of flexibility retrofitting. By 2025, the provincial grid will have a maximum chemical energy storage capacity of 7,000 MWh and a maximum new pumped storage capacity of 2,500 MW. This will allow up to 20,000 MW of thermal power units to participate in deep peaking, with a load factor as low as 30%.

[0184] The parameters of each power source, energy storage and time-of-use electricity price data are shown in Table 1 and Table 2 respectively. The benchmark on-grid electricity price of coal-fired power generation units (including desulfurization, denitrification and dust removal) is 0.3844 yuan / kWh. In 2022, the on-grid electricity price for newly approved onshore wind power projects, newly registered centralized photovoltaic power stations and industrial and commercial distributed photovoltaic projects will be implemented according to the local coal-fired power generation benchmark price. The compensation fee for deep peak regulation of thermal power generation is calculated according to the compensation standard of 0.3 yuan / kWh when the load is between 40% and 50% of the base load, and the compensation standard of 0.7 yuan / kWh when the load is between 30% and 40% of the base load. Currently, the commonly used chemical energy storages are 1h energy storage and 2h energy storage. The present invention selects 2h energy storage for optimization. This is because photovoltaics account for a high proportion of new energy in the province, and the time when absorption difficulties occur at noon is mostly 2-4 hours, and the unit cost of two-hour energy storage is lower.

[0185] Table 1 Peak shaving equipment parameters

[0186]

[0187] Table 2 Time-of-use electricity price list

[0188]

[0189] The present invention proposes a method for collaborative planning of deep peak regulation and composite energy storage for thermal power units. The implementation process is shown in the attached Figure 1 .

[0190] Step 1: Establish a multi-objective optimization model for the coordinated planning of deep peak regulation of thermal power units and composite energy storage with the best economy, the best carbon emission reduction and the least wind and solar power curtailment.

[0191] (1) Economically optimal model with maximum annual profit as the goal

[0192] max F1=Pr t +Pr n -C H -C s (1)

[0193] Where, Pr t and Pr nare thermal power income and new energy income respectively; C H with C s are the costs of pumped storage and chemical energy storage respectively; C d is the cost of curtailing wind and solar power.

[0194] The revenue from thermal power generation includes the revenue from the power grid company’s sales of thermal power, as well as the compensation paid by the power grid to the units participating in deep peak regulation.

[0195]

[0196] Where n is the total number of sampling points per year, and if one sampling point is taken every 15 minutes, n = 365 × 96 = 35040; i represents the i-th moment, the same below; P t is the actual output k of the thermal power unit s and k b are the grid electricity sales price and the thermal power on-grid benchmark price respectively; P td k is the power generated by the thermal power unit less than the lower limit of normal output (i.e. the power participating in deep peak regulation); td It is the compensation coefficient for deep peak-shaving electricity price of thermal power.

[0197] The income from new energy power generation is the income from electricity sales, after deducting the costs caused by abandoned wind and solar power.

[0198]

[0199] Where, P n is the total output of grid-connected new energy; P d is the abandoned wind and solar power; the revenue from selling electricity from new energy is the same as that from thermal power; k d is the cost coefficient for abandoned wind and solar power.

[0200] The cost of pumped-storage units includes the construction cost of new pumped-storage units and the annual operation and maintenance cost of all pumped-storage units.

[0201] C H =P H k H +P nH k nH / m H (4)

[0202] Where, P H and k H are the total installed capacity of the pumped storage units and their annual operation and maintenance cost coefficients; P nH and k nH are the newly added pumped storage power and the construction cost coefficient of the pumped storage unit; m H It is the service life of the pumped storage unit.

[0203] Assuming that all chemical energy storage is newly built, its cost includes new construction cost and operation and maintenance cost.

[0204]

[0205] Where, E es is the newly added chemical energy storage capacity; k es is the construction cost coefficient of chemical energy storage; m s and n s are the annual charge and discharge times of chemical energy storage and the total charge and discharge times in the entire life cycle; ν is the annual operation and maintenance cost coefficient of chemical energy storage.

[0206] (2) Maximum carbon emission reduction target

[0207] Carbon emission reductions are calculated by subtracting the incremental carbon emissions from deep peaking of coal-fired power plants from the reductions from combined energy storage. The carbon emission reductions from combined energy storage are the product of total power generation and the carbon emission coefficient. Since renewable energy output reduces the use of thermal power, the more renewable energy a system incorporates, the less thermal power output it generates. Therefore, when calculating the carbon emission reductions from combined energy storage, the carbon emission coefficient is calculated based on the carbon emission coefficient for conventional thermal power generation.

[0208] The objective function F2 is expressed as:

[0209] max F2=T p +T s -T t (6)

[0210] Where, T p and T s are the carbon emission reductions converted from the operation of pumped storage facilities and chemical energy storage respectively; T t The increase in carbon emissions caused by deep peak regulation of coal-fired power units.

[0211] The specific components are as follows:

[0212]

[0213] Where: k nf P is the carbon emission coefficient per unit power of the thermal power unit in normal operation; nf k is the output of thermal power units that have not entered deep peak regulation; df P is the unit carbon emission increment coefficient under deep peak regulation of thermal power units; p and P s are the real-time power of the pumped storage unit and the real-time power of chemical energy storage; μ gen and They are all 0-1 variables, with a value of 1 indicating that the pumped storage unit is generating electricity or the chemical energy storage is discharging.

[0214] (3) Minimum target for abandoned wind and solar power

[0215]

[0216] Constraints include system power balance constraints, equipment operation constraints, and decision variable constraints.

[0217] <1> System power balance constraints

[0218]

[0219] Where μ pum and μ gen are all 0-1 variables and satisfy μ pum μ gen =0, μ pum It is the pumping flag, when the value is 1, it is pumping; μ gen It is the power generation flag, when the value is 1, power generation occurs; and are all 0-1 variables and satisfy It is the chemical energy storage charging flag, when the value is 1, it is charging; It is the chemical energy storage discharge flag, when the value is 1, it is discharged; P L is the system load; P tie is the inter-provincial exchange power, with outflow being positive.

[0220] <2> Battery energy storage capacity balance constraints

[0221]

[0222] E s,min ≤E s (i)≤E s,max (11)

[0223] Where, E s (i) is the amount of electricity stored in the battery at time i; T s is the unit time; η c and η d are the charging and discharging efficiencies of battery energy storage respectively; since overcharging and over-discharging will affect the service life of chemical energy storage, E s,min With E s,max They are the lower and upper limits of battery energy storage capacity respectively.

[0224] <3> Battery energy storage output constraints

[0225]

[0226] Where, P sn The rated power of the battery energy storage is the rated power. The absorption and discharge power of the battery energy storage must be less than its rated power and meet the constraints of the remaining power in the battery.

[0227] <4> Pumped storage capacity constraints

[0228] E H,min ≤E H (i)≤E H,max (13)

[0229]

[0230] Where, E H (i) is the water storage capacity of the pumped storage power station at time i; E H,min and E H,max are the minimum and maximum water storage capacities of the pumped storage power station respectively; η pum and η gen are the pumping and power generation efficiencies, respectively.

[0231] <5> Output constraints of pumped storage units

[0232]

[0233] Step 2: Preprocess the input data based on the entropy weight-ideal method to obtain the comprehensive score of each decision variable, and then determine the order of inputting deep peak regulation of thermal power, pumped storage and chemical energy storage.

[0234] <1> Data normalization

[0235] In order to eliminate the impact of different dimensions on the evaluation results, it is necessary to normalize or standardize each indicator. The details are as follows:

[0236]

[0237] Where x ij is the jth objective function value of the i-th decision variable, p ij is x ij The normalized result of ; m and n represent the number of objective functions and the number of decision variables respectively.

[0238] <2> Calculate the information entropy e of the decision variable j , utility value d j and entropy weight w j , as follows:

[0239]

[0240] <3> Use ideal solution to determine the optimal solution

[0241] When using the ideal method to solve multi-objective decision-making problems, it is necessary to define a measure in the objective space to measure how close a solution is to the positive ideal solution and how far it is from the negative ideal solution. The core idea is to first select an ideal solution and a negative ideal solution, and then find the solution that is closest to the ideal solution and farthest from the negative ideal solution, which is the optimal solution.

[0242] For positive ideal solutions, we have:

[0243]

[0244] For negative ideal solutions, we have:

[0245]

[0246] <4> Normalized Matrix

[0247] z ij =x′ ij ×w j (20)

[0248]

[0249] <5> Calculate the best and worst values ​​and the best and worst distances

[0250]

[0251] Where z + is the optimal solution, z - is the worst solution, is the optimal distance, is the worst distance.

[0252] <6> Calculate weighted composite score

[0253]

[0254] Where C i It represents the comprehensive score of the i-th variable, with a value range of (0,1). The closer it is to 1, the closer the evaluation object is to the optimal level, and the higher the comprehensive score.

[0255] The genetic algorithm has a crossover probability of 0.9, a crossover distribution index of 20, a mutation probability of 0.1, a mutation distribution index of 20, a population size of 100, and an iteration number of 50. Based on the entropy weight method and the ideal solution, the optimal order of investment for each flexible resource is pumped storage → deep peak regulation with thermal power → chemical energy storage.

[0256] In step 3, the improved NSGA-II method is used to solve the model and obtain the Pareto optimal solution set. The comprehensive optimal solution is further obtained through fuzzy membership, and the deep peak regulation amount of the thermal power unit and the planned capacity of the new composite energy storage are obtained.

[0257] When improving the NSGA-II method for model solving, the improvements are mainly reflected in:

[0258] 1) The normal distribution crossover operator (NDX) is introduced to enhance the spatial search capability of the algorithm.

[0259] 2) An improved adaptive mutation adjustment method is proposed to increase the speed of population optimization. The traditional NSGA-II algorithm uses a polynomial mutation method. Because this mutation operator contains random and subjective parameters, it is highly random and converges slowly. The improved adaptive mutation adjustment method can achieve better convergence through its mechanism. It not only improves the convergence speed by utilizing the effect of mutation, but also enhances the diversity and stability of the population, thereby achieving a more optimal Pareto frontier distribution.

[0260] When solving the maximum value of the objective function, the fuzzy membership function of formula (24) is used to express it:

[0261]

[0262] When solving the objective function minimization problem, the fuzzy membership function of formula (25) is used to express it:

[0263]

[0264] Where, f j represents the j-th objective function value; and Represents the maximum and minimum values ​​of the jth objective function. For the multi-objective optimization problem in this paper, objective functions F1 and F2 are respectively designed to maximize the total system revenue and carbon emission reduction, so a maximum-type membership function is selected; objective function F3 is designed to minimize the amount of curtailed wind and solar power, so a minimum-type membership function is selected.

[0265] The satisfaction of all chromosomes in the Pareto optimal solution set is shown in formula (26).

[0266]

[0267] Where h represents overall satisfaction (standardized). A larger value indicates higher satisfaction.

[0268] The Pareto optimal solution set results of the coordinated planning of thermal power deep peak regulation and composite energy storage are attached. Figure 2 As shown in Table 3.

[0269] Table 3 Comprehensive optimal solution and single objective optimal solution

[0270]

[0271]

[0272] From the attached Figure 3It can be seen from Table 3 that: 1) The comprehensive optimal solution under multi-objective collaborative optimization has the highest satisfaction, with a value of 0.802. The collaborative optimization results are: the capacity of the thermal power unit transformation and the peak-shaving depth have reached the upper limit, which are 20,000MW and 30% respectively; the new pumped storage unit is 1,964MW, the new chemical energy storage is 403MWh, and the system's wind and solar curtailment rate is 0.09%. 2) When the economic optimum is taken as the goal, the optimization results are: the capacity of the thermal power unit transformation and the peak-shaving depth have reached the upper limit, the new pumped storage unit is 64MW, and the wind and solar curtailment rate is 0.273%. In this case, the system's wind and solar curtailment is large. This is because the cost of deep peak-shaving transformation of thermal power units is the lowest, so it is used first. The relationship between the scale of new pumped storage and the economy of the system can be seen in Figure 3 , it can be seen that the system's economic efficiency is optimal when the newly built pumped storage unit is 64MW. 3) If the only goal is to minimize carbon emissions, the optimization result is that the deep peak regulation of thermal power is zero, and the newly built chemical energy storage (70,000MWh) and pumped storage unit (2,000MW) both reach their maximum values. This is because deep peak regulation of thermal power increases the amount of coal burned per unit of power generation, leading to increased carbon emissions, while pumped storage and chemical energy storage can bring carbon emission reduction benefits. 4) If the only goal is to optimize the curtailment of wind and solar power, the optimization result is that all peak regulation methods reach their upper limits, at which point the system's curtailment of wind and solar power is minimized.

[0273] The iterative process of multi-objective optimization based on traditional NSGA-Ⅱ algorithm and improved NSGA-Ⅱ algorithm is shown in the attached Figure 3 . It can be seen that:

[0274] The improved NSGA-II algorithm achieves significantly faster iterative convergence than the traditional algorithm, reaching the optimal solution after just eight iterations, while the traditional NSGA-II algorithm requires 25 iterations. Because the optimal population size for convergence is the same, the two algorithms achieve the same optimal solution. This demonstrates that the improved NSGA-II algorithm improves both convergence speed and computational efficiency.

[0275] Example 2

[0276] The present invention proposes a device for collaborative planning of deep peak regulation and composite energy storage for thermal power units, comprising:

[0277] Acquisition unit: This unit acquires output data, load data, and external power data for wind power, photovoltaic power, and synchronous generators at 96 points per day, 365 days a year or multiple years. It also acquires planning data for renewable energy, pumped storage, and chemical energy storage in future years.

[0278] Modeling unit: Based on the three aspects of economy, carbon emission reduction and curtailed wind and solar power, a multi-objective planning model for deep peak regulation of thermal power units and composite energy storage coordination, as well as corresponding constraints, is established.

[0279] Solution unit: An improved NSGA-II algorithm based on the entropy weight ideal solution preprocesses data to solve the multi-objective planning model, obtaining the Pareto optimal solution set. The fuzzy membership is then used to obtain the comprehensive optimal solution, which in turn determines the deep peak regulation capacity of the thermal power units and the planned capacity of the newly built composite energy storage.

[0280] The description and application of the present invention here are illustrative and are not intended to limit the scope of the present invention to the above-mentioned embodiments. The relevant descriptions of the effects or advantages involved in the specification may not be reflected in the actual experimental examples due to the uncertainty of specific condition parameters or other factors, and the relevant descriptions of the effects or advantages are not used to limit the scope of the invention. Variations and changes to the embodiments disclosed here are possible, and the replacement of the embodiments and various equivalent components are well known to those of ordinary skill in the art. It should be clear to those skilled in the art that, without departing from the spirit or essential characteristics of the present invention, the present invention can be implemented in other forms, structures, arrangements, proportions, and with other components, materials and parts. Without departing from the scope and spirit of the present invention, other variations and changes can be made to the embodiments disclosed here.

Claims

1. A method for collaborative planning of deep peak regulation and composite energy storage for thermal power units, characterized in that: include: S1. Establish an economic model with the goal of maximizing annual revenue, a carbon emission reduction model with the goal of maximizing carbon emission reduction, and a wind and solar curtailment model with the goal of minimizing wind and solar curtailment. Based on the aforementioned economic model, carbon emission reduction model, and wind and solar curtailment model, establish a multi-objective optimization model for deep peak regulation of thermal power units and coordinated planning of composite energy storage. S2. Preprocessing input data based on the entropy weight-ideal method, where the input data is each decision variable, including thermal power deep peak regulation, pumped storage, and chemical energy storage, calculating a comprehensive score for each decision variable in the multi-objective optimization model, and determining the order of inputting thermal power deep peak regulation, pumped storage, and chemical energy storage according to the comprehensive score of each decision variable from largest to smallest; S3. An improved NSGA-II method is used to solve the model. The improved NSGA-II method generates offspring by introducing a normal distribution crossover operator and an adaptive adjustment mutation method based on the NSGA-II method, obtains a Pareto optimal solution set, and obtains a comprehensive optimal solution through a fuzzy membership function, thereby obtaining the deep peak-shaving transformation amount of the thermal power unit and the planned capacity of the newly built composite energy storage. The improved NSGA-II method is: (1) The initialization algorithm generates a parent population of n individuals, i.e., the newly created amount of each peak-shaving means, and calculates the objective function value according to the order of input of each peak-shaving means determined by data preprocessing; (2) Introducing the normal distribution crossover operator and the adaptive adjustment mutation method to perform crossover mutation to produce an offspring with n individuals; (3) Merge the parent and offspring populations, and perform non-dominated sorting and crowding calculation on the objective function value of the merged population; (4) Select the first n individuals to generate a new parent population; (5) Check whether the stopping condition is met, that is, all individuals are non-dominated solutions. If not, return to step (2). If so, output the Pareto optimal solution set. The comprehensive optimal solution obtained by the fuzzy membership function includes: The fuzzy membership function when solving the objective function maximization problem is: The fuzzy membership function when solving the objective function minimization problem is: Where, f j represents the j-th objective function value; and Indicates the maximum and minimum values ​​of the jth objective function; objective functions F1 and F2 are respectively the maximum total system benefit and the maximum carbon emission reduction, and the maximum membership function is selected; objective function F3 is the minimum amount of abandoned wind and solar power, and the minimum membership function is selected; The satisfaction of all individuals in the Pareto optimal solution set is: In the formula, h represents overall satisfaction, and a larger h value indicates a higher satisfaction.

2. The method for collaborative planning of deep peak regulation and composite energy storage for thermal power units according to claim 1, characterized in that: In S1, the economic model with the goal of maximizing annual revenue is established as follows: max F1=Pr t +Pr n -C H -C s Where, F1 is the annual income; Pr t and Pr n are thermal power generation income and new energy power generation income respectively; C H with C s are the cost of pumped storage units and chemical energy storage respectively; Among them, thermal power revenue Pr t This includes the power grid company's revenue from thermal power sales and the grid company's compensation to units participating in deep peak regulation, specifically: Where n is the total number of sampling points per year; i represents the i-th moment; P t is the actual output k of the thermal power unit s and k b are the grid electricity sales price and the thermal power on-grid benchmark price respectively; P td k is the power generated by the thermal power unit less than the lower limit of normal output, that is, the power participating in deep peak regulation; td is the compensation coefficient for deep peak-shaving electricity price of thermal power; New energy power generation income Pr n The revenue from electricity sales, after deducting the cost of abandoned wind and solar power, is as follows: Where, P n is the total output of grid-connected new energy; P d is the abandoned wind and solar power; k d is the cost coefficient of abandoned wind and solar power; Pumped storage unit cost C H This includes the construction cost of new pumped storage units and the annual operation and maintenance costs of all pumped storage units, specifically: C H =P H k H +P nH k nH / m H Where, P H and k H are the total installed capacity of the pumped storage units and their annual operation and maintenance cost coefficients; P nH and k nH are the newly added pumped storage power and the construction cost coefficient of the pumped storage unit; m H The service life of the pumped storage unit; Chemical energy storage cost C s Including new construction costs and operation and maintenance costs, specifically: Where, E es is the newly added chemical energy storage capacity; k es is the construction cost coefficient of chemical energy storage; m s and n s are the annual charge and discharge times of chemical energy storage and the total charge and discharge times in the entire life cycle; ν is the annual operation and maintenance cost coefficient of chemical energy storage.

3. The method for collaborative planning of deep peak regulation and composite energy storage for thermal power units according to claim 1, characterized in that: In S1, the carbon emission reduction model is established with the goal of maximizing carbon emission reduction: max F2=T p +T s -T t Where F2 is the carbon emission reduction, which is the carbon emission reduction brought by the composite energy storage minus the carbon emission increase brought by the deep peak regulation of the coal-fired power units; T p and T s are the carbon emission reductions converted from the operation of pumped storage units and chemical energy storage; T t The incremental carbon emissions from deep peak regulation of coal-fired power units; k nf P is the carbon emission coefficient per unit power of the thermal power unit in normal operation; nf k is the output of thermal power units that have not entered deep peak regulation; df P is the unit carbon emission increment coefficient under deep peak regulation of thermal power units; p and P s are the real-time power of the pumped storage unit and the real-time power of chemical energy storage respectively; P t is the actual output of the thermal power unit; μ gen and They are all 0-1 variables, and a value of 1 indicates that the pumped storage unit generates electricity and the chemical energy storage discharges.

4. The method for collaborative planning of deep peak regulation and composite energy storage for thermal power units according to claim 1, characterized in that: In S1, the wind power abandonment model is established with the goal of minimizing the amount of abandoned wind power as follows: Where F3 is the amount of wind and solar power curtailment, P d To abandon wind and solar power; Among them, the constraints of the wind and solar curtailment model include: 1) System power balance constraints: Where, i represents the i-th moment; P t is the actual output of the thermal power unit; P n is the total output of grid-connected new energy; μ pum and μ gen are all 0-1 variables and satisfy μ pum μ gen =0, μ pum is the pumping sign, μ pum =1 means pumping water, μ gen is the power generation symbol, μ gen =1 indicates power generation; and are all 0-1 variables and satisfy A symbol for charging chemical energy storage. When it is charging, It is the discharge mark of chemical energy storage. When it is discharged; P p and P s are the real-time power of the pumped storage unit and the real-time power of chemical energy storage respectively; P d P is the abandoned wind and solar power; L is the system load; P tie is the inter-provincial exchange power, with outflow being positive; 2) Battery energy storage power balance constraints: E s,min ≤E s (i)≤E s,max Where, E s (i) is the amount of electricity stored in the battery at time i; T s is the unit time; η c and η d are the charging and discharging efficiencies of battery energy storage; E s,min With E s,max They are the lower limit and upper limit of battery energy storage capacity respectively; 3) Battery energy storage output constraints: Where, P sn The rated power of the battery energy storage is 100%, and the absorption and discharge power of the battery energy storage must be less than its rated power; 4) Pumped storage unit capacity constraints: E H,min ≤E H (i)≤E H,max Where, E H (i) is the water storage capacity of the pumped storage power station at time i; E H,min and E H,max are the minimum and maximum water storage capacities of the pumped storage power station respectively; η pum and η gen are the pumping and power generation efficiencies, respectively; 5) Output constraints of pumped storage units:

5. The method for collaborative planning of deep peak regulation and composite energy storage for thermal power units according to claim 1, characterized in that: The specific steps of S2 include: S2-1. Normalize the economic target, carbon emission reduction target, or wind and solar curtailment target function value of each decision variable as follows: Where x ij is the jth objective function value of the i-th decision variable, p ij is x ij The normalized result of ; m and n represent the number of objective functions and the number of decision variables respectively; S2-2. Calculate the information entropy e of the objective function value of each decision variable j , utility value d j and entropy weight w j , specifically: S2-3. Use the ideal solution method to determine the optimal solution. For positive ideal solutions such as economic targets and carbon emission reduction targets, there are: For the negative ideal solution of the wind power curtailment target, we have: Where x j It refers to the set of the jth objective function values ​​of all decision variables, that is, x j ={x 1j ,x 2j ,x 3j }; The positive ideal solution is used to process x ij The larger the better the indicator, the negative ideal solution is used to process x ij The smaller the better the index, the better it is, and the larger the better the index x′ ij ; S2-4. Establish a normalized matrix: z ij =x′ ij ×w j Where x′ ij Refers to all ideal solutions, including positive ideal solutions and negative ideal solutions, Z is the comprehensive weighted index z ij The matrix formed; S2-5. Calculate the optimal solution, the worst solution, the optimal distance, and the worst distance of the i-th decision variable, specifically: Where z + is the optimal solution, z - is the worst solution, is the optimal distance, is the worst distance; S2-6. Calculate the weighted comprehensive score of the i-th variable, specifically: Where C i represents the comprehensive score of the i-th decision variable, with a value range of (0,1), C i The closer it is to 1, the closer the indicators of the decision variables are to the optimal level, and the higher the comprehensive score is.

6. A device for implementing the method for collaborative planning of deep peak regulation and composite energy storage for thermal power units according to any one of claims 1 to 5, characterized in that: include: The data acquisition unit is used to obtain the output data, load and external power data of wind power, photovoltaic power and synchronous generators in a certain area at the same time for one year or multiple years, and to obtain the planning data of new energy, pumped storage units and chemical energy storage in future years; A modeling unit is used to establish a multi-objective optimization model for deep peak regulation of thermal power units and coordinated planning of composite energy storage based on economic efficiency, carbon emission reduction, and curtailed wind and solar power, and to set corresponding constraints; The solving unit is used to solve the multi-objective optimization model using the improved NSGA-II algorithm that pre-processes data based on the entropy weight ideal solution, obtain the Pareto optimal solution set, and obtain the comprehensive optimal solution through fuzzy membership, thereby obtaining the deep peak regulation amount of the thermal power unit and the planned capacity of the newly built composite energy storage.

Images