A method and system for variable dimension optimization simulation of reservoir operation rules of a hydropower station

Abstract

This invention discloses a variable-dimensional optimization simulation method and system for hydropower station reservoir scheduling rules, belonging to the field of water resource efficiency utilization and reservoir scheduling technology. First, long-sequence operation data of the hydropower station reservoir are collected, and the optimal scheduling scheme is derived using dynamic programming. Simultaneously, grey relational analysis is used to determine the set of influencing factors X and the target factor Y, thereby constructing a sample set. Second, a nonlinear simulation model is constructed, and the optimal simulation model (Model) is determined using a variable-dimensional optimization method. * And its parameters. Finally, the newly obtained impact factor X new Input the best simulation model * The target variable Y is obtained to guide the operation and scheduling of the hydropower station reservoir. new This invention can fully utilize the reservoir's storage capacity to enhance water head efficiency. The system's total output and guarantee rate are superior to conventional methods such as scheduling charts. It has advantages such as good search performance and stability and reliability, and can provide important technical support for the scientific scheduling of large-scale reservoir groups.

Description

Technical Field

[0001] This invention belongs to the field of reservoir scheduling, and specifically relates to a method and system for variable-dimensional optimization simulation of reservoir scheduling rules for hydropower stations. Background Technology

[0002] Dispatch rules are an important technical means to guide the efficient operation of hydropower station reservoirs. However, power generation dispatch diagrams are mostly based on a few representative historical data, while the actual data and the historical data used in the scheme calculation are not entirely the same; moreover, existing methods cannot fully explore the long-term operation patterns, nor can they obtain valuable data experience from the historical dispatch process to guide the updating of dispatch rules, resulting in large deviations in conventional methods and significant room for improvement in practical applications. Summary of the Invention

[0003] The purpose of this invention is to provide a method and system for simulating variable-dimensional optimization of reservoir scheduling rules in hydropower stations. To achieve the above objective, this invention provides the following technical solution: a method for simulating variable-dimensional optimization of reservoir scheduling rules in hydropower stations, comprising the following steps:

[0004] S1. Based on the long-sequence data of the hydropower station reservoir within a preset time period, the optimal scheduling scheme and the optimal benefit value of the optimal scheduling scheme within the known time period are obtained by using dynamic programming.

[0005] S2. Use grey relational analysis to determine the set of influencing factors X and the target variable Y, and construct a set consisting of m samples.

[0006] S3. Based on the set of m samples, initialize the population, calculate the variables of each individual in the population, then correct the individual variables to the feasible region, and decode them to the corresponding model parameters.

[0007] S4. Construct nonlinear simulation model 1 and nonlinear simulation model 2, and then use nonlinear optimization methods to solve for the decision variables of nonlinear simulation model 1 and nonlinear simulation model 2 respectively. Then, obtain the intermediate variables of nonlinear simulation model 1 and nonlinear simulation model 2 respectively. Then, construct an optimized simulation model based on the intermediate variables and influencing factors, and input the influencing factors of m samples into the optimized simulation model in sequence to obtain the simulated values. Calculate the total deviation as the individual fitness of the population.

[0008] S5. Based on the fitness of each individual in the population, obtain the historical optimal parameters of each individual and the global optimal parameters of the population.

[0009] S6. The parameters of all individuals are updated using a dynamic search mechanism, and the better individuals are selected one by one by a greedy method based on the fitness of each individual.

[0010] S7. The parameters of all individuals are updated using a mapping learning mechanism, and the better individuals are selected one by one by a greedy method based on the fitness of each individual.

[0011] S8. Based on the maximum number of iterations preset during population initialization, determine whether the maximum number of iterations has been reached. If so, use the optimal parameters of the population obtained in the last iteration as the optimal parameters θ of the optimization simulation model. * And obtain the optimal simulation model corresponding to the optimal parameters as the best optimal simulation model. * Otherwise, proceed to step S5;

[0012] S9. The newly acquired impact factor X new Input the best optimization simulation model * The target variable Y is obtained to guide the operation and scheduling of the hydropower station reservoir. new .

[0013] Furthermore, step S1 described above specifically involves obtaining the optimal scheduling scheme and the optimal benefit value of the optimal scheduling scheme within the known time period using the following formula:

[0014]

[0015] Among them, V t V t-1 These represent the final and initial reservoir capacities of the hydropower station's reservoir during time period t, respectively; e t (V t V t-1 () represents the benefit value of the hydropower station reservoir during time period t; P represents the optimal benefit value of the hydropower station reservoir from time period t to the start of the scheduling period. t Δ t These represent the power output and constraint failure terms of the hydropower station reservoir during time period t, respectively; a t P is the number of hours in time period t; min Minimum output limit; b t This is for adjusting the coefficient.

[0016] Furthermore, in step S2 above, the influence factor is represented as a matrix: A = [X1, ..., X...] i ,…,X m ] T The target variable is represented as a column vector Y = [Y1, ..., Y2]. i ,…,Y m ] T The set of m samples constructed is: {(X1,Y1),…,(X... i ,Y i ),…,(X m ,Y m)};where X i Y i Let represent all influencing factors and the target variable of the i-th sample, respectively; m represents the number of samples; and T is the transpose symbol.

[0017] Furthermore, the aforementioned step S3 includes the following sub-steps:

[0018] S3.1 Initialize the population based on the set of m samples, and calculate the variables for each individual in the population as follows:

[0019] Z(k,r)=[Z(k,r,1),…,Z(k,r,l),…,Z(k,r,L)] T ; k∈[1,K]; r∈[1,R]; l∈[1,L]

[0020] Where Z(k,r) represents the r-th individual in the k-th iteration; Z(k,r,l) represents the l-th variable of the r-th individual in the k-th iteration; L is the number of variables; K is the maximum number of iterations; and R is the number of individuals.

[0021] S3.2. Adjust the individual variables to the feasible region according to the following formula, and decode them to the corresponding model parameters;

[0022]

[0023] Z(k,r,l)=min{max{Z min (l),w(k,r,l)},Z max (l)}

[0024] Where w(k,r,l) represents the l-th auxiliary variable of the r-th individual in the k-th iteration; Z max (l), Z min (l) represents the maximum and minimum values ​​of the l-th variable, respectively; r0 represents a random number uniformly distributed in the interval [0,1].

[0025] Furthermore, the aforementioned step S4 includes the following sub-steps:

[0026] S4.1 Construct nonlinear simulation model 1 and nonlinear simulation model 2 according to the following formulas, and solve for the decision variables a and γ of nonlinear simulation model 1 and nonlinear simulation model 2 respectively using nonlinear optimization methods:

[0027]

[0028]

[0029] Where, B = (H T H) -1 HT H = [K(A,A T [e], f = Y - ε1e, h = Y + ε2e; B represents the intermediate matrix; H represents the extended kernel matrix; C1 and C2 represent penalty coefficients; ε1 and ε2 are constant terms; e represents the unit column vector; K(·) represents the kernel function;

[0030] S4.2 Calculate the intermediate variables for nonlinear simulation model 1 and nonlinear simulation model 2 respectively using the following formulas:

[0031] [w1 b1] = B(fa),

[0032] [w2 b2] = B(h + γ);

[0033] S4.3 Construct the optimized simulation model according to the following formula:

[0034]

[0035] S4.4. Input the influence factors of m samples into the optimization simulation model f(X) sequentially according to the following formula to obtain the simulated values, and calculate the total deviation:

[0036]

[0037] Furthermore, in step S5 above, the historical optimal parameters for each individual and the global optimal parameters for the population are obtained according to the following formula:

[0038] PZ(k,r)=argmin{F[PZ(k-1,r)],F[Z(k,r)]},

[0039] GZ(k)=argmin{F[PZ(k,1)],…,F[PZ(k,I)],F[GZ(k-1)]};

[0040] Where PZ(k,r) represents the historical best parameter of the r-th individual in the k-th iteration; GZ(k) represents the historical best parameter of the population in the k-th iteration; F[GZ(k)] and F[PZ(k-1,r)] represent the fitness of GZ(k) and PZ(k-1,r), respectively.

[0041] Furthermore, in step S6 above, the parameters of all individuals are updated using a dynamic search mechanism according to the following formula, and based on the fitness of each individual, a greedy method is used to compare and select the better individual one by one:

[0042]

[0043] Z(k,r+1)=argmin{F[Z(k,r)],F[U(k,r)]};

[0044] Where levy(β) represents a Levy distribution random number; U(k,r) represents the r-th dynamically updated individual in the k-th iteration; F[U(k,r)] represents the fitness of U(k,r); r1 represents a random number uniformly distributed in the interval [0,1]; r2 represents a random number uniformly distributed in the interval [-1,1]; τ1 represents the dynamic search probability; and ρ1 represents the scaling factor.

[0045] Furthermore, in step S7 above, the parameters of all individuals are updated using a mapping learning mechanism according to the following formula, and based on the fitness of each individual, a greedy method is used to compare and select the better individual one by one:

[0046]

[0047] Z(k,r+1)=argmin{F[Z(k,r)],F[V(k,r)]},

[0048] Where r3 and r4 represent random numbers distributed in the interval [0,1]; τ2 represents the mapping learning probability; and V(k,r) represents the r-th mapping individual in the k-th iteration.

[0049] Furthermore, step S8 described above specifically involves: letting k = k + 1, and if the maximum number of iterations preset during population initialization is reached, then using GZ(k) obtained from the last iteration as the optimal parameter θ. * The corresponding model is denoted as the optimal simulation model. * .

[0050] Another aspect of the present invention proposes a variable-dimensional optimization simulation system for hydropower station reservoir scheduling rules, comprising: a preprocessing module, used to collect long-sequence data of hydropower station reservoirs within a preset time period, and use dynamic programming method to obtain the optimal scheduling scheme and the optimal benefit value of the optimal scheduling scheme within the known time period; and a sample construction module, used to use grey relational analysis method to determine the set of influencing factors X and the target variable Y, and construct a set consisting of m samples.

[0051] The simulation optimization module is used to obtain the optimal parameter θ. * and the best simulation model * ;

[0052] The generation and execution module is used to process the newly acquired impact factor X. new Input the best simulation model * The target variable Y is obtained to guide the operation and scheduling of the hydropower station reservoir. new .

[0053] Compared with existing technologies, this invention has the following advantages and beneficial effects: ① This invention uses a variable-dimensional optimization method to select simulation model calculation parameters, which can effectively improve convergence speed and search accuracy; ② This invention constructs a nonlinear optimization model based on kernel functions, which can map the original feature data to a high-dimensional space to identify its features and relationships, effectively enhancing the model's generalization ability; ③ This invention achieves deep integration of multiple new technologies and methods, which can obtain more significant comprehensive benefits than traditional methods such as power generation dispatching diagrams, effectively improving the utilization efficiency of hydropower station reservoir resources; ④ In practical work, this invention can dynamically update the simulation model using hydropower station reservoir operation data, and achieve online updates of dispatching rules by absorbing and refining dispatching experience knowledge, improving dispatching management decision-making capabilities and efficiency, and providing practical and innovative technical means for power generation dispatching of cascade reservoir groups. Attached Figure Description

[0054] Figure 1 This is a flowchart of the present invention;

[0055] Figure 2 This is a diagram showing the water level process of the hydropower station reservoir obtained by the present invention.

[0056] Figure 3 This is a diagram illustrating the power output process of the hydropower station reservoir obtained by the present invention.

[0057] Figure 4 This is a comparison chart of the results of the present invention and the traditional power generation dispatch chart. In the chart, (a) is a comparison chart of average power generation, (b) is a comparison chart of average power output, and (c) is a comparison chart of total water wastage. Detailed Implementation

[0058] To better understand the technical content of the present invention, specific embodiments are described below in conjunction with the accompanying drawings.

[0059] In this invention, various aspects of the invention are described with reference to the accompanying drawings, in which numerous illustrative embodiments are shown. Embodiments of the invention are not limited to those depicted in the drawings. It should be understood that the invention is implemented through any of the various concepts and embodiments described above, as well as the concepts and embodiments described in detail below, because the concepts and embodiments disclosed herein are not limited to any particular implementation. Furthermore, some aspects of the invention disclosed may be used alone or in any suitable combination with other aspects of the invention disclosed.

[0060] like Figure 1 The flowchart of this invention shown illustrates a variable-dimensional optimization simulation method for hydropower station reservoir scheduling rules, comprising the following steps:

[0061] S1. Based on the long-sequence data of the hydropower station reservoir within a preset time period, the optimal scheduling scheme and the optimal benefit value of the optimal scheduling scheme within the known time period are obtained by using dynamic programming.

[0062] Specifically, the optimal scheduling scheme and its optimal benefit value within the known time period are obtained using the following formula:

[0063]

[0064] Among them, V t V t-1 These represent the final and initial reservoir capacities of the hydropower station's reservoir during time period t, respectively; e t (V t V t-1 () represents the benefit value of the hydropower station reservoir during time period t; P represents the optimal benefit value of the hydropower station reservoir from time period t to the start of the scheduling period. t Δ t These represent the power output and constraint failure terms of the hydropower station reservoir during time period t, respectively; a t P is the number of hours in time period t; min Minimum output limit; b t This is for adjusting the coefficient.

[0065] S2. Using grey relational analysis, determine the set of influencing factors X and the target variable Y, and construct a set consisting of m samples; the influencing factors are represented as matrix A = [X1, ..., X2]. i ,…,X m ] T The target variable is represented as a column vector Y = [Y1, ..., Y2]. i ,…,Y m ] T The set of m samples constructed is: {(X1,Y1),…,(X... i ,Y i ),…,(X m ,Y m )};where X i Y i Let represent all influencing factors and the target variable of the i-th sample, respectively; m represents the number of samples; and T is the transpose symbol.

[0066] S3. Based on the set of m samples, initialize the population, calculate the variables for each individual in the population, then adjust the individual variables to the feasible region and decode them to the corresponding model parameters, including the following sub-steps S3.1 to S3.2:

[0067] S3.1 Initialize the population based on the set of m samples, and calculate the variables for each individual in the population as follows:

[0068] Z(k,r)=[Z(k,r,1),…,Z(k,r,l),…,Z(k,r,L)]T ; k∈[1,K]; r∈[1,R]; l∈[1,L]

[0069] Where Z(k,r) represents the r-th individual in the k-th iteration; Z(k,r,l) represents the l-th variable of the r-th individual in the k-th iteration; L is the number of variables; K is the maximum number of iterations; and R is the number of individuals.

[0070] S3.2. Adjust the individual variables to the feasible region according to the following formula, and decode them to the corresponding model parameters;

[0071]

[0072] Z(k,r,l)=min{max{Z min (l),w(k,r,l)},Z max (l)}

[0073] Where w(k,r,l) represents the l-th auxiliary variable of the r-th individual in the k-th iteration; Z max (l), Z min (l) represents the maximum and minimum values ​​of the l-th variable, respectively; r0 represents a random number uniformly distributed in the interval [0,1].

[0074] S4. Construct nonlinear simulation model 1 and nonlinear simulation model 2, and then use nonlinear optimization methods to solve for the decision variables of nonlinear simulation model 1 and nonlinear simulation model 2 respectively. Next, obtain the intermediate variables of nonlinear simulation model 1 and nonlinear simulation model 2 respectively. Then, construct an optimized simulation model based on the intermediate variables and influencing factors, and input the influencing factors of m samples into the optimized simulation model to obtain simulated values, and calculate the total deviation as the individual fitness of the population. Specifically, this includes the following steps S4.1 to S4.4: S4.1. Construct nonlinear simulation model 1 and nonlinear simulation model 2 respectively according to the following formulas, and use nonlinear optimization methods to solve for the decision variables a and γ of nonlinear simulation model 1 and nonlinear simulation model 2 respectively:

[0075]

[0076]

[0077] Where, B = (H T H) -1 H T H = [K(A,A T [e], f = Y - ε1e, h = Y + ε2e; B represents the intermediate matrix; H represents the extended kernel matrix; C1 and C2 represent penalty coefficients; ε1 and ε2 are constant terms; e represents the unit column vector; K(·) represents the kernel function;

[0078] S4.2 Calculate the intermediate variables for nonlinear simulation model 1 and nonlinear simulation model 2 respectively using the following formulas:

[0079] [w1 b1] = B(fa),

[0080] [w2 b2] = B(h + γ);

[0081] S4.3 Construct the optimized simulation model according to the following formula:

[0082]

[0083] S4.4. Input the influence factors of m samples into the optimization simulation model f(X) sequentially according to the following formula to obtain the simulated values, and calculate the total deviation:

[0084]

[0085] S5. Obtain the historical optimal parameters of each individual and the global optimal parameters of the population based on the fitness of each individual in the population; specifically, obtain the historical optimal parameters of each individual and the global optimal parameters of the population according to the following formula:

[0086] PZ(k,r)=argmin{F[PZ(k-1,r)],F[Z(k,r)]},

[0087] GZ(k)=argmin{F[PZ(k,1)],…,F[PZ(k,I)],F[GZ(k-1)]};

[0088] Where PZ(k,r) represents the historical best parameter of the r-th individual in the k-th iteration; GZ(k) represents the historical best parameter of the population in the k-th iteration; F[GZ(k)] and F[PZ(k-1,r)] represent the fitness of GZ(k) and PZ(k-1,r), respectively.

[0089] S6. Update the parameters of all individuals using a dynamic search mechanism, and select the better individual by comparing them one by one based on the fitness of each individual; specifically, update the parameters of all individuals using a dynamic search mechanism according to the following formula, and select the better individual by comparing them one by one based on the fitness of each individual:

[0090]

[0091] Z(k,r+1)=argmin{F[Z(k,r)],F[U(k,r)]};

[0092] Where levy(β) represents a Levy distribution random number; U(k,r) represents the r-th dynamically updated individual in the k-th iteration; F[U(k,r)] represents the fitness of U(k,r); r1 represents a random number uniformly distributed in the interval [0,1]; r2 represents a random number uniformly distributed in the interval [-1,1]; τ1 represents the dynamic search probability; and ρ1 represents the scaling factor.

[0093] S7. Update the parameters of all individuals using a mapping learning mechanism, and select the better individual by comparing them one by one based on the fitness of each individual; specifically, update the parameters of all individuals using a mapping learning mechanism according to the following formula, and select the better individual by comparing them one by one based on the fitness of each individual:

[0094]

[0095] Z(k,r+1)=argmin{F[Z(k,r)],F[V(k,r)]},

[0096] Where r3 and r4 represent random numbers distributed in the interval [0,1]; τ2 represents the mapping learning probability; and V(k,r) represents the r-th mapping individual in the k-th iteration.

[0097] S8. Based on the maximum number of iterations preset during population initialization, determine whether the maximum number of iterations has been reached. If so, use the optimal parameters of the population obtained in the last iteration as the optimal parameters θ of the optimization simulation model. * And obtain the optimal simulation model corresponding to the optimal parameters as the best optimal simulation model. * Otherwise, proceed to step S5;

[0098] The model parameters are determined using the variable-dimensional optimization method in steps S3 to S8.

[0099] S9. The newly acquired impact factor X new Input the best optimization simulation model * The target variable Y is obtained to guide the operation and scheduling of the hydropower station reservoir. new .

[0100] This invention also proposes a variable-dimensional optimization simulation system for hydropower station reservoir scheduling rules, comprising:

[0101] The preprocessing module collects long-sequence data of hydropower station reservoirs within a preset time period and uses dynamic programming to obtain the optimal scheduling scheme and its optimal benefit value within that known time period. The sample construction module uses grey relational analysis to determine the set of influencing factors X and the target variable Y, constructing a set of m samples.

[0102] The simulation optimization module is used to obtain the optimal parameter θ. * and the best simulation model * .

[0103] The generation and execution module is used to process the newly acquired impact factor X. new Input the best simulation model * The target variable Y is obtained to guide the operation and scheduling of the hydropower station reservoir. new .

[0104] To demonstrate the effects achieved by this invention, a reservoir of a leading hydroelectric power station in a certain river basin was selected for study. Figure 2 , Figure 3 This invention presents the hydropower station reservoir scheduling scheme. It can be seen that during long-term operation, the hydropower station experiences minimal power output disruption during pre-flood drawdown, flood season impoundment, and dry season compensation, with water levels and power output operating within preset ranges, fully demonstrating the rationality and practicality of the scheduling scheme. Figure 4 This paper compares the results of the present invention with those of traditional power generation dispatching diagrams. Figure (a) shows the average power generation comparison, (b) shows the average power output comparison, and (c) shows the total water wastage comparison. It can be seen that compared to traditional power generation dispatching diagrams, the present invention significantly increases average power generation and power output, while substantially reducing water wastage, effectively improving the overall operational efficiency of the hydropower station. In summary, the present invention can provide a stable and effective reservoir dispatching and operation scheme for hydropower stations.

[0105] While the present invention has been described above with reference to preferred embodiments, it is not intended to limit the invention. Those skilled in the art can make various modifications and refinements without departing from the spirit and scope of the invention. Therefore, the scope of protection of the present invention shall be determined by the claims.

Claims

1. A variable-dimensional optimization simulation method for reservoir scheduling rules in hydropower stations, characterized in that, Includes the following steps: S1. Based on the long-sequence data of the hydropower station reservoir within a preset time period, the dynamic programming method is used to obtain the optimal scheduling scheme within the known time period and the optimal benefit value of the optimal scheduling scheme. Step S1 specifically involves obtaining the optimal scheduling scheme and the optimal benefit value of the optimal scheduling scheme within the known time period using the following formula: ； in, , These represent the time periods of the hydropower station reservoir. The final storage capacity and the initial storage capacity; For the hydropower station reservoir during the period The benefit value; For the hydropower station reservoir from time period The optimal benefit value at the start of the scheduling period; , These represent the time periods of the hydropower station reservoir. Output and constraint failure items; For time period Hours; Minimum output limit; For adjustment coefficients; S2. Determine the set of influencing factors using grey relational analysis. Target variable , construct A set consisting of samples; S3, according to A set of samples is used to initialize the population, calculate the variables of each individual in the population, then correct the individual variables to the feasible region, and decode them to the corresponding model parameters. Step S3 includes the following sub-steps: S3.1 According to A population is initialized from a set of samples, and the variables of each individual in the population are calculated as follows: ； in, Indicates the first The second iteration Individual; Indicates the first The second iteration The first individual One variable; The number of variables; This represents the maximum number of iterations. The number of individuals; S3.

2. Adjust the individual variables to the feasible region according to the following formula, and decode them to the corresponding model parameters; ； ； in, Indicates the first The second iteration The first individual One auxiliary variable; , They represent the first The maximum and minimum values ​​of each variable; Represents a random number uniformly distributed in the interval [0,1]. S4. Construct nonlinear simulation model 1 and nonlinear simulation model 2, and then use nonlinear optimization methods to solve for the decision variables of nonlinear simulation model 1 and nonlinear simulation model 2 respectively. Then, obtain the intermediate variables of nonlinear simulation model 1 and nonlinear simulation model 2 respectively. Finally, construct an optimized simulation model based on the intermediate variables and influencing factors, and... The influencing factors of each sample are sequentially input into the optimization simulation model to obtain simulated values, and the total deviation is calculated as the individual fitness of the population. S5. Based on the fitness of each individual in the population, obtain the historical optimal parameters of each individual and the global optimal parameters of the population. S6. The parameters of all individuals are updated using a dynamic search mechanism, and the better individuals are selected one by one by a greedy method based on the fitness of each individual. S7. The parameters of all individuals are updated using a mapping learning mechanism, and the better individuals are selected one by one by a greedy method based on the fitness of each individual. S8. Based on the maximum number of iterations preset during population initialization, determine whether the maximum number of iterations has been reached. If so, use the optimal parameters of the population obtained in the last iteration as the optimal parameters for optimizing the simulation model. And obtain the optimal simulation model corresponding to the optimal parameters as the best optimal simulation model. Otherwise, proceed to step S5; S9. The newly acquired impact factors Input the optimal simulation model The target variables used to guide the scheduling and operation of hydropower station reservoirs were obtained. .

2. The variable-dimensional optimization simulation method for hydropower station reservoir scheduling rules according to claim 1, characterized in that, In step S2, the influence factors are represented as a matrix. The target variable is represented as a column vector. , constructed The set consisting of samples is: ;in, , They represent the first All influencing factors and target variables for each sample; Indicates the number of samples; This is the transpose symbol.

3. The variable-dimensional optimization simulation method for hydropower station reservoir scheduling rules according to claim 2, characterized in that, Step S4 includes the following sub-steps: S4.1 Construct nonlinear simulation model 1 and nonlinear simulation model 2 according to the following formulas, and solve for the decision variables of nonlinear simulation model 1 and nonlinear simulation model 2 respectively using nonlinear optimization methods. and : ； ； in, , , , ; Represents the intermediate matrix; Represents the extended kernel matrix; , Indicates the penalty coefficient; , For constant terms; Represents a unit column vector; Represents the kernel function; S4.2 Calculate the intermediate variables for nonlinear simulation model 1 and nonlinear simulation model 2 respectively using the following formulas: ； ； S4.3 Construct the optimized simulation model according to the following formula: ； S4.4, Apply the following formula: The influencing factors of each sample are sequentially input into the optimization simulation model. The simulated values ​​were obtained, and the total deviation was calculated: 。 4. The variable-dimensional optimization simulation method for hydropower station reservoir scheduling rules according to claim 3, characterized in that, In step S5, the historical optimal parameters for each individual and the global optimal parameters for the population are obtained using the following formula: ； ； in, Indicates the first The second iteration The historical optimal parameters for each individual; Indicates the first The historical optimal parameters of the population in the next iteration; , They represent , The degree of adaptability.

5. The variable-dimensional optimization simulation method for hydropower station reservoir scheduling rules according to claim 4, characterized in that, In step S6, the parameters of all individuals are updated using a dynamic search mechanism according to the following formula, and based on the fitness of each individual, a greedy method is used to compare and select the better individual one by one: ； ； in, Represents a Lévy distribution random number; Indicates the first The second iteration A dynamically updated individual; express The fitness of; Represents a random number uniformly distributed in the interval [0,1]. Represents a random number uniformly distributed in the interval [-1, 1]. Indicates the probability of dynamic search; Represents the scale factor.

6. The variable-dimensional optimization simulation method for hydropower station reservoir scheduling rules according to claim 5, characterized in that, In step S7, the parameters of all individuals are updated using the mapping learning mechanism according to the following formula, and based on the fitness of each individual, a greedy method is used to compare and select the better individual one by one: ； ； in, , Represents a random number distributed in the interval [0,1]. Represents the learning probability of the mapping; Indicates the first The second iteration One mapped individual.

7. The variable-dimensional optimization simulation method for hydropower station reservoir scheduling rules according to claim 4, characterized in that, Step S8 specifically involves: Let If the maximum number of iterations preset during population initialization is reached, then the result from the last iteration will be used. As the optimal parameter The corresponding model is denoted as the optimal simulation model. .

8. A variable-dimensional optimization simulation system for reservoir scheduling rules in hydropower stations, characterized in that, The system is used to perform the method according to any one of claims 1 to 7, the system comprising: The preprocessing module is used to collect long-sequence data of hydropower station reservoirs within a preset time period, and use dynamic programming methods to obtain the optimal scheduling scheme and the optimal benefit value of the optimal scheduling scheme within the known time period. The sample construction module is used to determine the set of influencing factors using the grey relational analysis method. Target variable , construct A set consisting of samples; The simulation optimization module is used to obtain the optimal parameters. and the best simulation model ; The generation and execution module is used to process the newly acquired impact factors. Input the best simulation model The target variables used to guide the scheduling and operation of hydropower station reservoirs were obtained. .

Images