A multi-objective two-stage flexible renovation planning method based on NSGA-II algorithm considering different renovation periods

Abstract

The present invention relates to the field of flexibility transformation planning of coal-fired units, and specifically to a multi-objective two-stage flexibility transformation planning method based on the NSGA-II algorithm taking into account different transformation periods, which specifically includes the following steps: obtaining the coal consumption parameters of each coal-fired unit, the expected parameter improvement of each unit after adopting different flexibility transformation technologies, the daily load and wind power load of the coal-fired unit operation; establishing an initialization multi-technology transformation plan based on NSGA-II in the planning stage, and establishing a mixed integer programming model for the operation of the coal-fired unit in the operation stage; solving the proposed multi-objective two-stage model according to the binary tournament selection method, the retention strategy, and the non-dominated rank sorting algorithm; and selecting the optimal solution according to the obtained Pareto solution set by the approximate ideal solution sorting method. The planning result obtained by the present invention can simultaneously coordinate the transformation cost target and the adjustment flexibility target, and through the refined modeling of the coal-fired unit operation, improves the planning equilibrium solution that conforms to the actual engineering practice.

Description

Technical Field

[0001] The present invention relates to the field of flexibility transformation planning for coal-fired units, and in particular to a multi-objective two-stage flexibility transformation planning method for coal-fired units based on the NSGA-II algorithm and considering different planning years. Background Art

[0002] Renewable energy sources, such as wind and photovoltaic power, are currently experiencing large-scale development, meaning that renewable power generation accounts for an increasing proportion of power system operations. At the same time, the intermittent and fluctuating nature of renewable energy generation places higher demands on the power system's regulatory capabilities. Coal-fired units, the primary source of power generation in the current power system, have proven cost-effective in retrofitting their flexibility to improve their regulatory capabilities. Therefore, rationally planning the timing and technology selection for these retrofits is a key issue in promoting power system transformation.

[0003] The objective function of the flexibility transformation planning problem of coal-fired units mainly considers two aspects: one is to set a reasonable objective function to reasonably plan the coal-fired units that need to be transformed, and the other is to reasonably model the flexibility transformation technology.

[0004] For the flexibility retrofit planning of coal-fired power plants, currently established planning models are mostly based on a single economic objective function. To assess the improved regulation capacity after the flexibility retrofit, a wind curtailment penalty coefficient or load shedding penalty coefficient is typically set in the objective function to convert this into wind curtailment costs or load shedding as part of the economic objective model. However, during this conversion process, the wind curtailment penalty coefficient or load shedding penalty coefficient is manually set based on various model parameters. This makes the objective function susceptible to the coefficient settings, resulting in a non-optimal compromise in the planning results. Therefore, planning results derived from a single economic objective have limitations. Multi-objective optimization techniques are key to solving optimization problems involving multiple objective functions. With recent advances in engineering technology, the NSGA-II algorithm, as a classic multi-objective optimization algorithm, has attracted the attention of researchers in various engineering fields. It is currently widely used in power system planning for energy storage optimization and unit commitment economic dispatch. However, research on multi-objective solution algorithms for the flexibility retrofit planning of coal-fired power plants is relatively limited.

[0005] Regarding the modeling of flexibility retrofit technologies, it is currently recognized that the ability of flexibility retrofits to improve coal-fired units primarily includes three performance indicators: minimum technical output, ramp rate, and start-up time. Because minimum technical output directly reflects a coal-fired unit's ability to absorb renewable energy generation, some current research literature only considers reducing minimum technical output as a basis for planning flexibility retrofits. Other literature comprehensively considers minimum technical output, ramp rate, and start-up and shutdown times, but fails to consider the diversity of retrofit technologies. In other words, few current retrofit technologies can simultaneously address all three flexibility characteristics, resulting in planning results that are overly optimistic compared to actual results. Furthermore, in actual operation, the start-up and shutdown processes of coal-fired units are very complex, from unit startup to the unit entering a dispatchable state. Therefore, existing references pay little attention to the rapid start-up and shutdown of coal-fired units.

[0006] Based on the above reasons, it is necessary to propose a new coal-fired unit planning technology to finely model the flexibility transformation technology of coal-fired units and achieve the coordination of multiple optimization objectives. Summary of the Invention

[0007] In order to overcome the above-mentioned technical problems, the purpose of the present invention is to propose a flexible transformation planning model for coal-fired units taking into account multi-objective optimization, and to solve the model based on the NSGA-II algorithm. The model generates a solution that meets the constraints through a repair strategy in the planning stage, and evaluates the benefits of the transformation planning results through the operation of the established coal-fired units in the operation stage. Finally, the Pareto solution set obtained by the NSGA-II algorithm is optimized by the approximate ideal solution sorting method. This invention avoids the problem of unbalanced planning results caused by the artificial setting of wind curtailment and load shedding penalty coefficients, and the problem of planning results not being in line with the actual project due to imprecise modeling of flexibility transformation technology.

[0008] The purpose of the present invention can be achieved through the following technical solutions:

[0009] The multi-objective two-stage planning method for the flexibility transformation of coal-fired units includes the following steps:

[0010] Step 1: Obtain the per-unit values ​​of the typical daily load curve and wind power curve for the power system to be transformed during the planning period, the relevant operating parameters of the coal-fired units included in the system, the optional flexibility transformation technologies for each coal-fired unit, and the corresponding expected parameter improvements;

[0011] Step 2: Based on the parameters obtained in Step 1, a multi-objective two-stage planning model is established. This model is divided into a planning stage and an operation stage. The objectives are to minimize the sum of the transformation planning cost and the operating cost and to minimize the regulation power loss, respectively. In the planning stage, investment constraints and parameter limit constraints are established to obtain the investment cost. In the operation stage, a mixed integer model for the typical daily operation of the coal-fired unit is established with the objective of minimizing the power regulation loss, taking into account power limit constraints, ramping constraints, and minimum start-stop time constraints, to obtain the power regulation loss.

[0012] Step 3: Perform iterative optimization using the NSGA-II algorithm. Initialize the population strategy to generate a combination of solutions that meet the constraints of the planning phase. Based on the obtained objective function value, perform non-dominated ranking, crossover, and mutation operations. After a finite number of iterations, obtain the Pareto solution set of the model.

[0013] Step 4: Use the approximate ideal solution sorting method to select the optimal solution in the Pareto solution set described in step 3;

[0014] The step 1 is specifically as follows: establishing a system planning period set Ω in units of years. Y , y is the annual number in the planning period; the 24-hour period in a typical day is the scheduling period, and the set of scheduling periods in the entire typical day is Ω T , t is the dispatch period number, obtain the load data per unit value D of the power system in period t on a typical day y in the planning year y,t , wind power data W y,t ;

[0015] Obtain various parameters of the coal-fired units contained in the system, where the set of coal-fired units is Ω G , the unit number is g, the maximum technical output of coal-fired unit g is The original minimum output is The original climbing rate is R g , start time is ST g For coal-fired unit g, the set of optional flexibility transformation technologies is Ω g,H , h is the number corresponding to the technology set, and the minimum technology load that can be reduced by technology h is The climbing rate that can be increased is ΔR g,h , the start-up time can be reduced to ΔST g,h ;

[0016] Furthermore, the step 2 is specifically as follows:

[0017] A1. Define the planning stage to minimize the transformation investment cost within the planning period as the objective function F1, as shown in (1) and (2)

[0018]

[0019]

[0020] Where U g,y,h is the binary decision variable of coal-fired unit g for technology h in planning year y. g,y,h =1 means the decision is effective otherwise U g,y,h =0, The average annual transformation cost of using transformation technology h is: is the transformation cost per unit capacity corresponding to technology h, ρ is the discount rate, and r is the expected operating life of the coal-fired unit;

[0021] Establish planning stage constraints, including technical constraints and operating parameter constraints. Technical constraints are shown in (3) and (4).

[0022]

[0023] The operating parameter constraints are shown in (5) and (6)

[0024]

[0025] In the formula Indicates the minimum technical output that can be reduced by coal-fired unit g, ST g It represents the time required for coal-fired unit g to start up and enter the dispatchable state;

[0026] A2. Define the operation phase to minimize the power regulation loss during the planning period as the objective function F2, as shown in (7)

[0027]

[0028] In the formula represents the installed capacity of wind power in year y, It represents the per-unit value of wind power curtailment during period t on a typical day in planning year y. represents the load capacity in year y, It represents the per-unit load shedding power value in time period t on a typical day in the planning year y;

[0029] Establish the constraints of the operation phase, including operation state constraints, dispatchable state constraints, start-stop constraints, coal-fired unit power output constraints, ramp constraints and power balance constraints. The operation state constraints are shown in (8)

[0030] I g,y,t -I g,y,t-1 =γ g,y,t -η g,y,t (8)

[0031] When I g,y,t=1 means that the coal-fired unit g is in operation during the typical day t period of the planning year y, otherwise it is 0, γ g,y,t =1,η g,y,t =1 indicates that the coal-fired unit g executes the start and stop instructions in the corresponding period, otherwise it is 0;

[0032] The schedulable state variables are shown in (9)

[0033] φ g,y,t -φ g,y,t-1 =ν g,y,t -η g,y,t (9)

[0034] Where φ g,y,t =1 indicates that the coal-fired unit g is in a dispatchable state during the typical day t period of the planning year y, otherwise it is 0, ν g,y,t =1 indicates that the coal-fired unit g begins to enter the dispatchable stage during the corresponding period, otherwise it is 0;

[0035] The start-stop state constraints are shown in (10)-(13):

[0036]

[0037]

[0038] In the formula They represent the minimum continuous start-up and shutdown time of coal-fired unit g, t′ and d represent the scheduling period number different from t. It represents the startup time of coal-fired unit g in planning year y after the combination of planning layer transformation schemes;

[0039] The power output constraints of coal-fired units are shown in (14)-(16):

[0040]

[0041] Where P g,y,t represents the output power of coal-fired unit g during period t on a typical day in planning year y, SU g , SD g They represent the starting and stopping power of coal-fired unit g, It represents the minimum technical output of coal-fired unit g in planning year y after the combination of planning layer transformation schemes;

[0042] The ramping constraints of coal-fired units are shown in (17)-(19)

[0043]

[0044] In the formula It represents the corresponding ramp rate of coal-fired unit g in planning year y after the combination of planning layer transformation schemes;

[0045] The power balance constraints are shown in (20)-(23):

[0046]

[0047]

[0048] Where ε wind , ε load Represent the annual growth rates of wind power installed capacity and load capacity respectively.

[0049] Furthermore, the step 3 is specifically as follows:

[0050] B1. Generate a combination of solutions that meet the constraints of the planning stage by initializing the population strategy. The specific steps are to construct integer chromosomes and randomly generate N G ×N Y uniform integer array, where N G 、N Y are the number of coal-fired units in the system under study and the number of planned years, and then the elements that do not meet the constraints are repaired through the repair strategy; for coal-fired unit g, it is determined whether there are repeated elements in the planned years. If so, only the first repeated element is retained, and then the elements that do not meet the constraints are repaired in descending order. ΔST g,h Ω g,H The optional technologies are arranged as follows Determine in turn whether the generated solution satisfies the formulas (5) and (6). If not, or Select an element with a smaller value than the element that violates the constraint as a replacement, and then re-check whether the constraint is satisfied. If not, repeat the above steps until the generated planning scheme no longer violates the constraint. At this time, the generated population is P0, and the population size is N size , calculate the objective function value of individual P0, perform non-dominated rank sorting and crowding calculation as the initial reference solution;

[0051] B2. Obtain the parent population Q through binary tournament selection k , the crossover and inheritance probabilities are p c 、p m , through crossover, mutation operator and repair strategy, we get the offspring population S k , and obtain the next generation population P k =S k ∪Q k ;

[0052] B3. Perform non-dominated rank sorting and congestion calculation based on the objective function value described in B1. Let k = 1 be the number of iterations of the current algorithm, and let the non-dominated rank Rank in the kth iteration bek Satisfy Rank k,1 ≤Rank k ≤Rank k,max , Rank k,1 Rank k,max They represent the first non-dominated rank and the maximum non-dominated rank in the kth iteration, respectively, and retain P k Middle front N size individuals with a small distance between their non-dominated ranking and crowding degree;

[0053] B4. Determine whether k reaches the set maximum number of iterations K max , if k <K max Then calculate k=k+1 and return to B2. If yes, terminate the loop and get the population P. k , select non-dominated rank k =1 as the Pareto solution set;

[0054] Furthermore, the step 4 is specifically as follows:

[0055] C1. According to the Pareto solution set, the jth objective function F′ of the i-th individual in the solution set i,j Calculate normalization, the calculation method is shown in (24):

[0056]

[0057] Where n i,j is the element of the standardized decision matrix, Ω Pareto is the set of individuals in the Pareto solution set, Ω J is the set composed of objective functions, i and j are the labels of Pareto solution set and objective function set respectively;

[0058] C2. Calculate the weighted matrix, where the element of the jth target of the i-th individual in the matrix is ​​x i,j =ω j n i,j ,ω j is the target weight of the jth objective function, and

[0059] C3. Determine the ideal solution A + and non-ideal solution A - , the calculation method is shown in (25) and (26):

[0060]

[0061] In the formula In turn, they represent the optimal solution, the worst solution of the first objective function and the optimal solution, the worst solution of the second objective function;

[0062] C4. Calculate Ω in sequence Pareto Ideal solution of individual distance A + and non-ideal solution A - distance As shown in (27) and (28):

[0063]

[0064] C5. Calculate the relative distance between ideal points Then the optimal solution i in the Pareto solution set is best The method to determine

[0065]

[0066] Where i best Represents Ω Pareto Thus, the Pareto optimal solution of the multi-objective two-stage planning method for the flexibility transformation of coal-fired units is obtained.

[0067] Compared with other existing achievements, the beneficial effects of the present invention include: (1) The technology proposed in the present invention avoids the imbalance of planning results caused by artificially setting the penalty coefficients for wind curtailment and load shedding, and can achieve the coordinated compromise of the model through the NSGA-II optimization algorithm and planning model solving technology; (2) The technology proposed in the present invention actually considers the technical impact of flexibility transformation technology on coal-fired units, and finely models the startup and operation process of coal-fired units, thereby improving the objectivity of planning results and making them more in line with engineering reality. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] The present invention will be further described below with reference to the accompanying drawings.

[0069] Figure 1 It is a schematic flow chart of the method of the present invention. DETAILED DESCRIPTION

[0070] The following will be combined with the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0071] See also Figure 1 The multi-objective two-stage planning method for flexibility transformation of coal-fired units of the present invention comprises the following steps:

[0072] Step 1: Obtain the per-unit values ​​of the typical daily load curve and wind power curve for the power system to be transformed during the planning period, the relevant operating parameters of the coal-fired units included in the system, the optional flexibility transformation technologies for each coal-fired unit, and the corresponding expected parameter improvements. The detailed process is as follows:

[0073] Establish the system planning period set Ω in years Y , y is the annual number in the planning period; the 24-hour period in a typical day is the scheduling period, and the set of scheduling periods in the entire typical day is Ω T , t is the dispatch period number, obtain the load data per unit value D of the power system in period t on a typical day y in the planning year y,t , wind power data W y,t ;

[0074] Obtain various parameters of the coal-fired units contained in the system, where the set of coal-fired units is Ω G , the unit number is g, the maximum technical output of coal-fired unit g is The original minimum output is The original climbing rate is R g , start time is ST g For coal-fired unit g, the set of optional flexibility transformation technologies is Ω g,H , h is the number corresponding to the technology set, and the minimum technology load that can be reduced by technology h is The climbing rate that can be increased is ΔR g,h , the start-up time can be reduced to ΔST g,h ;

[0075] Step 2: Based on the parameters obtained in Step 1, a multi-objective two-stage planning model is established. This model is divided into a planning stage and an operation stage. The objectives are to minimize the sum of the transformation planning cost and the operating cost and to minimize the regulation power loss, respectively. In the planning stage, investment constraints and parameter limit constraints are established to obtain the investment cost. In the operation stage, a mixed integer model for the typical daily operation of the coal-fired unit is established with the goal of minimizing the power regulation loss, taking into account the power limit constraint, ramping constraint, and minimum start-stop time constraint, to obtain the power regulation loss. The detailed process is as follows:

[0076] A1. Define the planning stage to minimize the transformation investment cost within the planning period as the objective function F1, as shown in (1) and (2)

[0077]

[0078] Where U g,y,h is the binary decision variable of coal-fired unit g for technology h in planning year y. g,y,h =1 means the decision is effective otherwise The average annual transformation cost of using transformation technology h is: is the transformation cost per unit capacity corresponding to technology h, ρ is the discount rate, and r is the expected operating life of the coal-fired unit;

[0079] Establish planning stage constraints, including technical constraints and operating parameter constraints. Technical constraints are shown in (3) and (4).

[0080]

[0081] The operating parameter constraints are shown in (5) and (6)

[0082]

[0083] In the formula Indicates the minimum technical output that can be reduced by coal-fired unit g, ST g It represents the time required for coal-fired unit g to start up and enter the dispatchable state;

[0084] A2. Define the operation phase to minimize the power regulation loss during the planning period as the objective function F2, as shown in (7)

[0085]

[0086] In the formula represents the installed capacity of wind power in year y, It represents the per-unit value of wind power curtailment during period t on a typical day in planning year y. represents the load capacity in year y, It represents the per-unit load shedding power value in time period t on a typical day in the planning year y;

[0087] Establish the constraints of the operation phase, including operation state constraints, dispatchable state constraints, start-stop constraints, coal-fired unit power output constraints, ramp constraints and power balance constraints. The operation state constraints are shown in (8)

[0088] I g,y,t -I g,y,t-1 =γ g,y,t -η g,y,t (8)

[0089] When I g,y,t =1 means that the coal-fired unit g is in operation during the typical day t period of the planning year y, otherwise it is 0, γ g,y,t =1,η g,y,t =1 indicates that the coal-fired unit g executes the start and stop instructions in the corresponding period, otherwise it is 0;

[0090] The schedulable state variables are shown in (9)

[0091] φ g,y,t -φ g,y,t-1=ν g,y,t -η g,y,t (9)

[0092] Where φ g,y,t =1 indicates that the coal-fired unit g is in a dispatchable state during the typical day t period of the planning year y, otherwise it is 0, ν g,y,t =1 indicates that the coal-fired unit g begins to enter the dispatchable stage during the corresponding period, otherwise it is 0;

[0093] The start-stop state constraints are shown in (10)-(13):

[0094]

[0095] In the formula They represent the minimum continuous start-up and shutdown time of coal-fired unit g, t′ and d represent the scheduling period number different from t. It represents the startup time of coal-fired unit g in planning year y after the combination of planning layer transformation schemes;

[0096] The power output constraints of coal-fired units are shown in (14)-(16):

[0097]

[0098]

[0099] Where P g,y,t represents the output power of coal-fired unit g during period t on a typical day in planning year y, SU g , SD g They represent the starting and stopping power of coal-fired unit g, It represents the minimum technical output of coal-fired unit g in planning year y after the combination of planning layer transformation schemes;

[0100] The ramping constraints of coal-fired units are shown in (17)-(19):

[0101]

[0102] In the formula It represents the corresponding ramp rate of coal-fired unit g in planning year y after the combination of planning layer transformation schemes;

[0103] The power balance constraints are shown in (20)-(23):

[0104]

[0105] Where ε wind , ε load Represent the annual growth rates of wind power installed capacity and load capacity respectively.

[0106] Step 3: Perform iterative optimization using the NSGA-II algorithm. Initialize the population strategy to generate a combination of solutions that meet the constraints of the planning phase. Based on the obtained objective function value, perform fast non-dominated sorting, crossover, mutation, and other operations. After a finite number of iterations, obtain the Pareto solution set of the model. The detailed process is as follows:

[0107] B1. Generate a combination of solutions that meet the constraints of the planning stage by initializing the population strategy. The specific steps are to construct integer chromosomes and randomly generate N G ×N Y uniform integer array, where N G 、N Y are the number of coal-fired units in the system under study and the number of planned years, and then the elements that do not meet the constraints are repaired through the repair strategy; for coal-fired unit g, it is determined whether there are repeated elements in the planned years. If so, only the first repeated element is retained, and then the elements that do not meet the constraints are repaired in descending order. ΔST g,h Ω g,H The optional technologies are arranged as follows Determine in turn whether the generated solution satisfies the formulas (5) and (6). If not, or Select an element with a smaller value than the element that violates the constraint as a replacement, and then re-check whether the constraint is satisfied. If not, repeat the above steps until the generated planning scheme no longer violates the constraint. At this time, the generated population is P0, and the population size is N size , calculate the objective function value of individual P0, perform non-dominated rank sorting and crowding calculation as the initial reference solution;

[0108] B2. Obtain the parent population Q through binary tournament selection k , the crossover and inheritance probabilities are p c 、p m , through crossover, mutation operator and repair strategy, we get the offspring population S k , and obtain the next generation population P k =S k ∪Q k ;

[0109] B3. Perform non-dominated rank sorting and congestion calculation based on the objective function value described in B1. Let k = 1 be the number of iterations of the current algorithm, and let the non-dominated rank Rank in the kth iteration be k Satisfy Rank k,1 ≤Rank k ≤Rank k,max , Rank k,1 Rank k,max They represent the first non-dominated rank and the maximum non-dominated rank in the kth iteration, respectively, and retain Pk Middle front N size individuals with a small distance between their non-dominated ranking and crowding degree;

[0110] B4. Determine whether k reaches the set maximum number of iterations K max , if k <K max Then calculate k=k+1 and return to B2. If yes, terminate the loop and get the population P. k , select non-dominated rank k =1 as the Pareto solution set;

[0111] Step 4: Use the approximate ideal solution sorting method to select the optimal solution in the Pareto solution set described in step 3. The detailed process is as follows:

[0112] C1. According to the Pareto solution set, the jth objective function F′ of the i-th individual in the solution set i,j Calculate normalization, the calculation method is shown in (24):

[0113]

[0114] Where n i,j is the element of the standardized decision matrix, Ω Pareto is the set of individuals in the Pareto solution set, Ω J is the set composed of objective functions, i and j are the labels of Pareto solution set and objective function set respectively;

[0115] C2. Calculate the weighted matrix, where the element of the jth target of the i-th individual in the matrix is ​​x i,j =ω j n i,j ,ω j is the target weight of the jth objective function, and

[0116] C3. Determine the ideal solution A + and non-ideal solution A - , the calculation method is shown in (25) and (26):

[0117]

[0118] In the formula In turn, they represent the optimal solution, the worst solution of the first objective function and the optimal solution, the worst solution of the second objective function;

[0119] C4. Calculate Ω in sequence Pareto Ideal solution of individual distance A + and non-ideal solution A - distance As shown in (27) and (28)

[0120]

[0121] C5. Calculate the relative distance between ideal points Then the optimal solution i in the Pareto solution set is best The method to determine

[0122]

[0123] Where i best Represents Ω Pareto Thus, the Pareto optimal solution of the multi-objective two-stage planning method for the flexibility transformation of coal-fired units is obtained.

[0124] The above contents are merely examples and explanations of the present invention. Those skilled in the art may make various modifications or additions to the described specific embodiments or replace them in similar ways. As long as they do not deviate from the invention or exceed the scope defined in this specification, they should all fall within the scope of protection of the present invention.

Claims

1. A multi-objective two-stage planning method for flexibility transformation of coal-fired units, characterized by: The steps include: Step 1: Obtain the per-unit values ​​of the typical daily load curve and wind power curve for the power system to be transformed during the planning period, the relevant operating parameters of the coal-fired units included in the system, the optional flexibility transformation technologies for each coal-fired unit, and the corresponding expected parameter improvements; Step 2: Based on the parameters obtained in Step 1, a multi-objective two-stage planning model is established. This model is divided into a planning stage and an operation stage. The objectives are to minimize the sum of the transformation planning cost and the operating cost and to minimize the regulation power loss, respectively. In the planning stage, investment constraints and parameter limit constraints are established to obtain the investment cost. In the operation stage, a mixed integer model for the typical daily operation of the coal-fired unit is established with the objective of minimizing the power regulation loss, taking into account the power limit constraint, ramping constraint, and minimum start-stop time constraint, to obtain the power regulation loss. Step 3: Perform iterative optimization using the NSGA-II algorithm. Initialize the population strategy to generate a combination of solutions that meet the constraints of the planning phase. Based on the obtained objective function value, perform non-dominated ranking, crossover, and mutation operations. After a finite number of iterations, obtain the Pareto solution set of the model. Step 4: Use the approximate ideal solution sorting method to select the optimal solution in the Pareto solution set described in step 3; The step 1 is specifically as follows: Establish a systematic planning period set in years , y It is the annual number within the planning period; The scheduling period is 24 hours in a typical day, and the set of scheduling periods in the entire typical day is , t The dispatch period number is used to obtain the power system y Typical day t Load data per unit value for the period , wind power data ; Get the parameters of the coal-fired units contained in the system, where the coal-fired units are composed of , the unit number is g , coal-fired units g The maximum technical output is , the original minimum output is , the original climbing rate is , start time is For coal-fired units The flexible transformation technology set available is , h is the number corresponding to the technology set, technology The corresponding minimum technical load that can be reduced is , the climbing rate that can be improved is , the startup time can be reduced to ; The step 2 is specifically as follows: A1. Define the planning phase with the objective function of minimizing the renovation investment cost within the planning period. F 1, as shown in (1) and (2) (1) (2) In the formula For coal-fired units g In the planning year y Technology h The binary decision variable is Indicates that the decision takes effect otherwise , Adopting transformation technology h The average annual renovation cost is for h The transformation cost per unit capacity of the technology, ρ is the discount rate, r is the expected operating life of the coal-fired unit; Establish planning stage constraints, including technical constraints and operating parameter constraints. Technical constraints are shown in (3) and (4). (3) (4) The operating parameter constraints are shown in (5) and (6) (5) (6) In the formula Coal-fired units g The minimum technical output that can be reduced. Coal-fired units g The time required from startup to entering the dispatchable state; A2. Define the operation phase with the objective function of minimizing power regulation loss during the planning period F 2, as shown in (7) (7) In the formula represents the installed capacity of wind power in year y, Indicates planning year y typical day t The per-unit value of wind power curtailment during the period, Indicates the y Annual load capacity, Indicates planning year y typical day t per-unit value of load shedding power in the time period; Establish the operation phase constraints, including operation state constraints, dispatchable state constraints, start-stop constraints, coal-fired unit power output constraints, ramp constraints and power balance constraints. The operation state constraints are shown in (8) (8) in style I g,y,t =1 indicates the planning year y typical day t Period coal-fired units g In running state, otherwise 0, γ g,y,t =1, η g,y,t =1 indicates that the coal-fired unit g executes the start and stop instructions in the corresponding period, otherwise it is 0; The schedulable state variables are shown in (9) (9) In the formula ϕ g,y,t =1 indicates the planning year y typical day t Period coal-fired units g In the schedulable state, otherwise it is 0, ν g,y,t =1 indicates coal-fired units in the corresponding period g Start entering the schedulable stage, otherwise it is 0; The start-stop state constraints are shown in (10)-(13): (10) (11) (12) (13) In the formula 、 Represents coal-fired units g Minimum continuous start-up and shutdown time, 、 d Both indicate that they are different from t The scheduling period number, Coal-fired units g In the planning year y The corresponding startup time after the combination of planning-level transformation solutions; The power output constraints of coal-fired units are shown in (14)-(16): (14) (15) (16) In the formula P g,y,t Indicates planning year y typical day t Period coal-fired units g The output power, SU g 、 SD g Represents coal-fired units g Starting and stopping power, Coal-fired units g In the planning year y The minimum technical output corresponding to the combination of planning-level transformation solutions; The ramping constraints of coal-fired units are shown in (17)-(19): (17) (18) (19) In the formula Coal-fired units g In the planning year y The corresponding climbing rate after the combination of planning layer transformation schemes; The power balance constraints are shown in (20)-(23): (20) (21) (22) (23) In the formula ε wind , ε load Represent the annual growth rates of wind power installed capacity and load capacity respectively.

2. The multi-objective two-stage planning method for flexibility transformation of coal-fired units according to claim 1 is characterized in that: The step 3 is specifically as follows: B1. Generate a combination of solutions that meet the constraints of the planning stage by initializing the population strategy. The specific steps are to construct integer chromosomes and randomly generate uniform integer array, where N G 、 N Y are the number of coal-fired units in the system under study and the number of planned years, and then the elements that do not meet the constraints are repaired through the repair strategy; for coal-fired units g , determine whether there are repeated elements in the planning year, if so, only retain the first repeated element, and then sort them in descending order 、 Ω g,H The optional technologies are arranged as follows 、 , judge whether the generated solution satisfies the formula (5) and (6) in turn. If not, or Select an element with a smaller value than the element that violates the constraint as a replacement, and then re-check whether the constraint is satisfied. If not, repeat the above steps until the generated planning scheme no longer violates the constraint. At this time, the generated population is P 0, the population size is N size ,calculate P The objective function value of individual 0 is used to perform non-dominated ranking and crowding calculation as the initial reference solution; B2. Obtaining the parent population through binary tournament selection Q k , crossover and inheritance probabilities are p c 、 p m , the offspring population is obtained through crossover, mutation operator and repair strategy S k , and obtain the next generation population ; B3. Perform non-dominated ranking and congestion calculation based on the objective function value described in B1. Let k =1 is the number of iterations of the current algorithm, let k The non-dominated level in iteration satisfy , 、 Respectively represent k The first non-dominated rank and the maximum non-dominated rank in the iteration are retained. P k Center front N size individuals with a small distance between their non-dominated ranking and crowding degree; B4. Judgment k Whether the maximum number of iterations set has been reached K max ,like k < K max Then calculate k = k +1 and return to B2, if yes, terminate the loop and get the population P k , select the non-dominated level All individuals of are regarded as Pareto solution sets.

3. The multi-objective two-stage planning method for flexibility transformation of coal-fired units according to claim 2 is characterized in that: The step 4 is specifically as follows: C1. According to the Pareto solution set, the first i The individual's j objective function Calculate normalization, the calculation method is shown in (24): (24) In the formula n i,j is the element of the standardized decision matrix, Ω Pareto is the set of individuals in the Pareto solution set, Ω J is the set of objective functions, i 、 j are the labels of the Pareto solution set and the objective function set respectively; C2. Calculate the weighted matrix, the first i Individual j The elements of the target are , ω j For the j The objective weight of the objective function, and ; C3. Determine the ideal solution A + and non-ideal solutions A - , the calculation method is shown in (25) and (26): (25) (26) In the formula 、 、 、 In turn, they represent the optimal solution, the worst solution of the first objective function and the optimal solution, the worst solution of the second objective function; C4. Calculate Ω in sequence Pareto Ideal solution of individual distance A + and non-ideal solutions A - distance 、 , as shown in (27) and (28): (27) (28) C5. Calculate the relative distance between ideal points , then the optimal solution in the Pareto solution set is i best The determination method is: (29) In the formula i best Represents Ω Pareto Thus, the Pareto optimal solution of the multi-objective two-stage planning method for the flexibility transformation of coal-fired units is obtained.

Images