Cascade power station scheduling optimization method and system considering risk and benefit game equilibrium

Abstract

The present invention discloses a method and system for optimizing cascade power station scheduling that considers risk and benefit game equilibrium, relating to the technical field of power system scheduling. The key points of the technical solution are: establishing a multi-task scheduling model; establishing a risk and benefit game model, and converting the Pareto efficiency equilibrium strategy solution problem of the cooperative game between the two entities of risk and benefit in the risk and benefit game model into a two-objective optimization solution problem; using the Pareto efficiency equilibrium strategy set as the initial scheduling strategy; inputting the initial scheduling strategy into a scheduling simulation model and collecting scheduling simulation information after running it; evaluating and analyzing the scheduling simulation information based on an external utility function, and selecting the initial scheduling strategy with the best external utility value from the Pareto non-inferior solutions as the actual scheduling strategy. The present invention converts the multi-objective optimization problem into a scheduling risk and benefit game problem, achieving dimensionality reduction of the multi-objective optimization problem.

Description

Technical Field

[0001] The present invention relates to the technical field of power system dispatching, and more specifically, to a cascade power station dispatching optimization method and system considering risk and benefit game equilibrium. Background Art

[0002] In addition to completing designated load planning tasks, daily scheduling models must also balance multiple tasks based on the specific seasonal characteristics of the schedule. Multi-objective balancing issues are primarily addressed through multi-objective modeling and the use of heuristic algorithms such as target dimensionality reduction, particle swarm optimization, and genetic algorithms to find Pareto-non-inferior solutions. Target dimensionality reduction can involve weighting targets or converting them into constraints.

[0003] However, existing solutions to multi-objective equilibrium problems have the following problems: 1. If objective dimensionality reduction is adopted, the objective weights or the allowable value ranges must be given in advance. In actual operation, this is too subjective and leads to a large degree of arbitrariness. 2. When using heuristic algorithms to solve the set of non-inferior solutions for multi-objective optimization, the solutions given are still multiple solutions. In actual use, the problem of how to select one solution from multiple sets of solutions to execute is still faced, and how to make this selection is still under research. 3. When using heuristic algorithms, when there are more than two objectives, the above algorithms are less efficient. 4. The above scheduling problem does not consider the risk of the scheduling strategy itself, but instead regards risk as a variable with a value within a certain range. Within this range, risks are treated equally. For example, in practice, even if the water level in front of the dam is given a range of values, the model may prefer a higher or lower water level during optimization. In this case, the risk of flood control during the flood season caused by a higher water level is obviously higher than that caused by a lower water level. 5. Current scheduling models do not consider the impact of decisions on the future.

[0004] Therefore, how to research and design a cascade power station scheduling optimization method and system that can overcome the above-mentioned defects and consider the risk and benefit game equilibrium is a problem that we urgently need to solve. Summary of the Invention

[0005] In order to address the deficiencies in the prior art, the purpose of the present invention is to provide a cascade power station scheduling optimization method and system that takes into account the risk and benefit game equilibrium, introduces risk and benefit decision-making methods into scheduling, and converts the multi-objective scheduling problem into a benefit-risk cooperative game problem, so that the cascade power station scheduling optimization is more in line with reality, and the accuracy and rationality of cascade power station scheduling are better.

[0006] The above technical objectives of the present invention are achieved through the following technical solutions:

[0007] First, a method for optimizing the dispatch of cascade power plants considering the game equilibrium of risk and benefit is provided, including the following steps:

[0008] Establish a multi-task scheduling model within the corresponding time period based on multiple dam statuses, power load plans, future water inflow forecasts, and time factors;

[0009] According to the scheduling task types in the multi-task scheduling model, a corresponding risk and benefit game model is established, and the Pareto efficiency equilibrium strategy solution problem of the risk and benefit two-subject cooperative game in the risk and benefit game model is converted into a two-objective optimization solution problem.

[0010] The NSGA-III algorithm is used to solve the Pareto efficiency equilibrium strategy set of the game, and the Pareto efficiency equilibrium strategy set is used as the initial scheduling strategy;

[0011] Establish a corresponding scheduling simulation model according to the scheduling scenario, input the initial scheduling strategy into the scheduling simulation model and collect the scheduling simulation information after running it;

[0012] The scheduling simulation information is evaluated and analyzed based on the pre-built external utility function, and the initial scheduling strategy with the best external utility value is selected from the Pareto non-inferior solutions as the actual scheduling strategy.

[0013] Furthermore, the scheduling scenarios are divided into flood season scheduling and dry season scheduling based on time characteristics.

[0014] Furthermore, the contradiction between risks and benefits of flood season scheduling is the contradiction between water storage and flood control.

[0015] Furthermore, the contradiction between the risks and benefits of the dry season scheduling is the contradiction between the water consumption rate and the risk of water withdrawal.

[0016] Furthermore, the risk and benefit game model corresponding to the flood season scheduling is specifically as follows:

[0017] For flood season benefits, under a given power generation load plan, the smaller the sum of the total outflow of the cascade power station group, the greater the flood season benefits;

[0018] For flood season risk parties, the safety redundancy of the water level in front of the dam of the cascade hydropower station group is used as a measure. The more the water level is below the safety level, the greater the safety redundancy, the smaller the risk, and the greater the benefit to the risk party.

[0019] Furthermore, the risk and benefit game model corresponding to the dry season scheduling is specifically as follows:

[0020] In terms of dry season benefits, the lower the comprehensive water consumption rate of the cascade power station group, the better. The comprehensive water consumption rate is obtained by dividing the total power generation water of all power stations by the total power generation.

[0021] For the dry season risk aspect, the difference between the planned water level in front of the dam around the power station and the target water level is defined as the measure of the dry season risk. The greater the deviation, the greater the risk.

[0022] Furthermore, the objective function of the two-objective optimization problem corresponding to the flood season scheduling is specifically:

[0023]

[0024]

[0025]

[0026] Among them, g1 represents the first optimization goal of flood season scheduling, Indicates the minimum value of multiple calculated values ​​generated by selecting a set of data s1, q1, s2, q2, s3, q3 as input. It represents the maximum value among the multiple minimum values ​​obtained by inputting multiple sets of different data s1, q1, s2, q2, s3, q3; g2 represents the second optimization goal of flood season scheduling, Indicates the minimum value of multiple calculated values ​​obtained by selecting multiple sets of different data s1, q1, s2, q2, s3, q3 as input; st represents the constraint condition, C1-C12 represent different constraint conditions; i represents the power station number, i value 1 represents the first power station at the upstream of the cascade power station group, i value 2 represents the second power station at the midstream of the cascade power station group, i value 3 represents the third power station at the downstream of the cascade power station group; q i represents the power generation flow of the i-th power station; h i represents the water level in front of the dam of the i-th power station; s i represents the abandoned water flow of the i-th power station, p i represents the output of the i-th power station; s1 represents the abandoned water flow of the first-level power station, p1 represents the output of the first-level power station; q1 represents the power generation flow of the first-level power station; s2 represents the abandoned water flow of the second-level power station; p2 represents the output of the second-level power station; q2 represents the power generation flow of the second-level power station; s3 represents the abandoned water flow of the third-level power station; p3 represents the output of the third-level power station; q3 represents the power generation flow of the third-level power station; ds i represents the adjusted abandoned water flow of the i-th power station; H i represents the lowest water level in front of the dam of the i-th power station; represents the highest water level of the i-th power station; q i represents the minimum power generation flow of the i-th level power station; represents the maximum power generation flow of the i-th power station; s i It represents the minimum water discharge when the gate of the i-th power station is opened; Q i represents the minimum outflow rate of the i-th level power station; represents the maximum outflow of the i-th power station; Pdaily_plan_pwr Indicates the total planned output of the cascade power station group; nhq i represents the output function of the i-th power station; I1 represents the total inflow of the first-level power station; I j A represents the total inflow of the j-th power station; j represents the natural inflow to the reservoir of the j-th power station; s j-1 represents the abandoned water flow of the j-1th power station; q f-1 V1 represents the power generation flow of the j-1th power station; 0 Represents the initial storage capacity of the first-stage power station; Vx 1 End-of-period storage capacity of the first-stage power station; represents the initial storage capacity of the j-th power station; represents the final storage capacity of the j-th power station; s j represents the abandoned water flow of the j-th power station; q j represents the power generation flow of the j-th power station; b i represents the tailwater level of the i-th power station; f i 1 represents the function for calculating the water level in front of the dam based on the storage capacity of the i-th power station, and the function is fitted based on the experimental data of the i-th power station; f i 2 It represents the function for calculating the tailwater level based on the power generation flow and the abandoned water flow, and is obtained by fitting the experimental data of power station i.

[0027] Furthermore, the objective function of the two-objective optimization problem corresponding to the dry season scheduling is specifically:

[0028]

[0029]

[0030]

[0031] Among them, g3 represents the first optimization goal of dry season scheduling, Indicates the minimum value of multiple calculated values ​​generated by selecting a set of data s1, q1, s2, q2, s3, q3 as input. It represents the maximum value among multiple minimum values ​​obtained by inputting multiple sets of different data s1, q1, s2, q2, s3, q3; g4 represents the second optimization goal of dry season scheduling, Indicates the minimum value of multiple calculated values ​​obtained by selecting multiple sets of different data s1, q1, s2, q2, s3, q3 as input; st represents the constraint condition, C1-C12 represent different constraint conditions; i represents the power station number, i value 1 represents the first power station at the upstream of the cascade power station group, i value 2 represents the second power station at the midstream of the cascade power station group, i value 3 represents the third power station at the downstream of the cascade power station group; q i represents the power generation flow of the i-th power station; H i，weeklt represents the target water level in front of the dam of the i-th power station; h i represents the water level in front of the dam of the i-th power station; f i It represents the function of calculating the final water level in front of the dam of the reservoir of the i-th power station through the initial water level, power generation flow, abandoned water flow and inflow; s i represents the abandoned water flow of the i-th power station; p i represents the output of the i-th power station; s1 represents the abandoned water flow of the first-level power station, p1 represents the output of the first-level power station; q1 represents the power generation flow of the first-level power station; s2 represents the abandoned water flow of the second-level power station; p2 represents the output of the second-level power station; q2 represents the power generation flow of the second-level power station; s3 represents the abandoned water flow of the third-level power station; p3 represents the output of the third-level power station; q3 represents the power generation flow of the third-level power station; ds i represents the adjusted abandoned water flow of the i-th power station; H i represents the lowest water level in front of the dam of the i-th power station; represents the highest water level of the i-th power station; q i represents the minimum power generation flow of the i-th level power station; represents the maximum power generation flow of the i-th power station; s i It represents the minimum water discharge when the gate of the i-th power station is opened; Q i represents the minimum outflow rate of the i-th level power station; represents the maximum outflow of the i-th power station; P daily_plan_pwr Indicates the total planned output of the cascade power station group; nhq i represents the output function of the i-th power station; I1 represents the total inflow of the first-level power station; I j A represents the total inflow of the j-th power station; j represents the natural inflow to the reservoir of the j-th power station; s j-1 represents the abandoned water flow of the j-1th power station; q j-1 V1 represents the power generation flow of the j-1th power station; 0 Represents the initial storage capacity of the first-stage power station; V1 1 End-of-period storage capacity of the first-stage power station; represents the initial storage capacity of the j-th power station; represents the final storage capacity of the j-th power station; s j represents the abandoned water flow of the j-th power station; q j represents the power generation flow of the j-th power station; b i represents the tailwater level of the i-th power station; f i 1 represents the function for calculating the water level in front of the dam based on the storage capacity of the i-th power station, and the function is fitted based on the experimental data of the i-th power station; f i 2 It represents the function for calculating the tailwater level based on the power generation flow and the abandoned water flow, and is obtained by fitting the experimental data of power station i.

[0032] Furthermore, the external utility function is constructed by using the future basin water inflow probability forecast and future load forecast, and adopting the Monte Carlo simulation method to estimate the impact of current decisions on the achievement of future long-term goals, thereby constructing the external utility function.

[0033] Secondly, a cascade power station dispatch optimization system that considers the risk and benefit game equilibrium is provided, including:

[0034] The scheduling establishment module is used to establish a multi-task scheduling model within the corresponding time period based on multiple dam states, power load plans, future water inflow forecasts and time factors;

[0035] The game conversion module is used to establish a corresponding risk and benefit game model according to the scheduling task type in the multi-task scheduling model, and convert the Pareto efficiency equilibrium strategy solution problem of the risk and benefit two-subject cooperative game in the risk and benefit game model into a two-objective optimization solution problem;

[0036] The game solving module is used to solve the Pareto efficiency equilibrium strategy set of the game using the NSGA-III algorithm and use the Pareto efficiency equilibrium strategy set as the initial scheduling strategy;

[0037] The scheduling simulation module is used to establish a corresponding scheduling simulation model according to the scheduling scenario, input the initial scheduling strategy into the scheduling simulation model and collect scheduling simulation information after running;

[0038] The evaluation and analysis module is used to evaluate and analyze the scheduling simulation information based on the pre-built external utility function, and select the initial scheduling strategy with the best external utility value from the Pareto non-inferior solutions as the actual scheduling strategy.

[0039] Compared with the prior art, the present invention has the following beneficial effects:

[0040] 1. The cascade power station scheduling optimization method proposed in this invention, which considers the risk and benefit game balance, converts the multi-objective optimization problem into a scheduling risk and benefit game problem, thus achieving dimensionality reduction of the multi-objective optimization problem.

[0041] 2. The present invention proposes a cascade power station scheduling optimization method that considers the balance of risk and benefit game. It introduces the mutual game of risk and benefit into the scheduling process, and the resulting scheduling strategy has Pareto optimality.

[0042] 3. The present invention introduces external utility functions, future basin inflow forecasts and future load forecast information, and adopts Monte Carlo simulation methods to estimate the impact of game equilibrium strategy decisions on future scheduling, solving the quantitative evaluation of the future impact of the strategy. The resulting scheduling strategy has the characteristic of optimal future impact benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The drawings described herein are used to provide a further understanding of the embodiments of the present invention, constitute a part of this application, and do not constitute a limitation of the embodiments of the present invention. In the drawings:

[0044] Figure 1 is a flow chart in an embodiment of the present invention;

[0045] Figure 2 It is a system block diagram in an embodiment of the present invention. DETAILED DESCRIPTION

[0046] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with examples and drawings. The exemplary embodiments of the present invention and their descriptions are only used to explain the present invention and are not intended to limit the present invention.

[0047] Example 1: Cascade power station scheduling optimization method considering risk and benefit game equilibrium, such as Figure 1 As shown, the following steps are included:

[0048] S1: A multi-task scheduling model is established for the corresponding time period based on multiple dam states, power load plans, future water inflow forecasts, and time factors. The multi-task scheduling model is established using existing technologies and will not be described in detail here.

[0049] S2: Establish a corresponding risk and benefit game model based on the scheduling task types in the multi-task scheduling model, and transform the Pareto efficiency equilibrium strategy solution problem of the risk and benefit two-subject cooperative game in the risk and benefit game model into a two-objective optimization solution problem; the scheduling task types are divided into risk management tasks and benefit improvement tasks;

[0050] S3: Use the NSGA-III algorithm to solve the Pareto efficiency equilibrium strategy set of the game and use the Pareto efficiency equilibrium strategy set as the initial scheduling strategy. NSGA is an improved genetic algorithm, which has been improved to the third generation, namely NSGA-III.

[0051] S4: Establish a corresponding scheduling simulation model according to the scheduling scenario, input the initial scheduling strategy into the scheduling simulation model and collect scheduling simulation information after running it;

[0052] S5: Evaluate and analyze the scheduling simulation information based on the pre-built external utility function, and select the initial scheduling strategy with the best external utility value from the Pareto non-inferior solutions as the actual scheduling strategy.

[0053] Scheduling scenarios are divided into flood season scheduling and dry season scheduling based on temporal characteristics. The risk-benefit contradiction in flood season scheduling is the conflict between water storage and flood control. The risk-benefit contradiction in dry season scheduling is the conflict between water consumption rate and drawdown risk.

[0054] Specifically, the risk and benefit game model corresponding to flood season scheduling is as follows: for the flood season benefit side, under a given power generation load plan, the smaller the sum of the total outflow of the cascade power station group, the greater the flood season benefit side's benefits; for the flood season risk side, the safety redundancy of the dam front water level of the cascade power station group is used as the measure, the more it is below the safety water level, the greater the safety redundancy, the smaller the risk, and the greater the benefit of the risk side.

[0055] The risk and benefit game model corresponding to dry season scheduling is specifically as follows: on the dry season benefit side, the lower the comprehensive water consumption rate of the cascade power station group, the better. The comprehensive water consumption rate is obtained by dividing the total power generation water of all power stations by the total power generation; on the dry season risk side, the difference from the planned water level target in front of the dam around the power station is defined as the measure of dry season risk. The greater the deviation, the greater the risk.

[0056] The action sets of the benefit party and the risk party are consistent. In the process of the game, it is assumed that the average of the actions of the two is used as the system execution action. In the actual process, because the risk party and the benefit party belong to the same system, that is, they belong to the same group, it is assumed that the group will give certain rewards and punishments to the benefit party and the risk party based on the comprehensive evaluation of the end-of-period scheduling, that is, the decision-makers have common interests. Therefore, the problem can be converted into a cooperative game problem, and the benefits of the entire system can be maximized through the cooperation of the risk decision-makers and the benefit decision-makers.

[0057] In the cooperative game between two game players, the game equilibrium must be Pareto optimal, because if it is not, it means that there is room for further improvement for one party without harming the other party. Whether it is the improvement of the risk side or the benefit side, it means that the overall benefit of the system has been improved, that is, either the risk is reduced relative to before while the benefit remains unchanged; or the benefit is improved relative to before while the risk remains unchanged; or both are improved; therefore, the cooperative game between two game players can be converted into a multi-objective optimization problem, but it should be noted that if there are more than two game players, there will be "small groups", and at this time the cooperative game problem may not be able to be converted into a multi-objective optimization problem.

[0058] Specifically, the objective function of the two-objective optimization problem corresponding to flood season scheduling is:

[0059]

[0060]

[0061]

[0062] Among them, g1 represents the first optimization goal of flood season scheduling, Indicates the minimum value of multiple calculated values ​​generated by selecting a set of data s1, q1, s2, q2, s3, q3 as input. It represents the maximum value among the multiple minimum values ​​obtained by inputting multiple sets of different data s1, q1, s2, q2, s3, q3; g2 represents the second optimization goal of flood season scheduling, Indicates the minimum value of multiple calculated values ​​obtained by selecting multiple sets of different data s1, q1, s2, q2, s3, q3 as input; st represents the constraint condition, C1-C12 represent different constraint conditions; i represents the power station number, i value 1 represents the first power station at the upstream of the cascade power station group, i value 2 represents the second power station at the midstream of the cascade power station group, i value 3 represents the third power station at the downstream of the cascade power station group; q i represents the power generation flow of the i-th power station; h i represents the water level in front of the dam of the i-th power station; s i represents the abandoned water flow of the i-th power station, p i represents the output of the i-th power station; s1 represents the abandoned water flow of the first-level power station, p1 represents the output of the first-level power station; q1 represents the power generation flow of the first-level power station; s2 represents the abandoned water flow of the second-level power station; p2 represents the output of the second-level power station; q2 represents the power generation flow of the second-level power station; s3 represents the abandoned water flow of the third-level power station; p3 represents the output of the third-level power station; q3 represents the power generation flow of the third-level power station; ds i represents the adjusted abandoned water flow of the i-th power station;H i represents the lowest water level in front of the dam of the i-th power station; represents the highest water level of the i-th power station; q i represents the minimum power generation flow of the i-th level power station; represents the maximum power generation flow of the i-th power station; s i It represents the minimum water discharge when the gate of the i-th power station is opened; Q i represents the minimum outflow rate of the i-th level power station; represents the maximum outflow of the i-th power station; P daily_plan_pwr Indicates the total planned output of the cascade power station group; nhq i represents the output function of the i-th power station; I1 represents the total inflow of the first-level power station; I j A represents the total inflow of the j-th power station; j represents the natural inflow to the reservoir of the j-th power station; s j-1 represents the abandoned water flow of the j-1th power station; q j-1 V1 represents the power generation flow of the j-1th power station; 0 Represents the initial storage capacity of the first-stage power station; V1 1 End-of-period storage capacity of the first-stage power station; represents the initial storage capacity of the j-th power station; represents the final storage capacity of the j-th power station; s j represents the abandoned water flow of the j-th power station; q j represents the power generation flow of the j-th power station; b i represents the tailwater level of the i-th power station; f i 1 represents the function for calculating the water level in front of the dam based on the storage capacity of the i-th power station, and the function is fitted based on the experimental data of the i-th power station; f i 2 It represents the function for calculating the tailwater level based on the power generation flow and the abandoned water flow, and is obtained by fitting the experimental data of power station i.

[0063] Specifically, g1 is the first optimization target of flood season scheduling, which represents the risk index of the flood season. The closer the water level is to the maximum water level allowed by the power station, the greater the risk. According to the "barrel effect principle", the water level risk of multiple power stations depends on the risk of the power station with the greatest risk, that is, the power station whose water level is closest to the maximum water level of the hydropower station, i.e. The lowest risk is then used as the dispatching target, thereby reducing the shortcomings of risk management in the cascade power station group.

[0064] g2 represents the efficiency indicator during the flood season. During this period, water is abundant, and water abandonment often occurs. When the future total output plan is given, how to reduce water abandonment is the primary efficiency indicator pursued by the power station. Therefore, minimizing the total water abandonment of the three power stations is set as the efficiency optimization goal.

[0065] Each set of input data s1, q1, s2, q2, s3, q3 can get multiple sets of h i In flood season scheduling, first solve the different groups of the same set of input data. h i The difference of data, select the minimum value of multiple differences as the solution target; then solve the maximum value of the minimum values ​​selected by different groups of input data as the solution target. The above is the first target optimization solution problem of flood season scheduling. In addition, each group of input data s1, q1, s2, q2, s3, q3 is used in the input calculation process as ds i ,q i The minimum sum is taken as the second objective optimization solution problem, and finally the two-objective optimization solution problem corresponding to flood season scheduling can be achieved by combining the constraints.

[0066] In addition, the objective function of the two-objective optimization problem corresponding to dry season scheduling is specifically:

[0067]

[0068]

[0069]

[0070] Among them, g3 represents the first optimization goal of dry season scheduling, Indicates the minimum value of multiple calculated values ​​generated by selecting a set of data s1, q1, s2, q2, s3, q3 as input. It represents the maximum value among multiple minimum values ​​obtained by inputting multiple sets of different data s1, q1, s2, q2, s3, q3; g4 represents the second optimization goal of dry season scheduling, Indicates the minimum value of multiple calculated values ​​obtained by selecting multiple sets of different data s1, q1, s2, q2, s3, q3 as input; st represents the constraint condition, C1-C12 represent different constraint conditions; i represents the power station number, i value 1 represents the first power station at the upstream of the cascade power station group, i value 2 represents the second power station at the midstream of the cascade power station group, i value 3 represents the third power station at the downstream of the cascade power station group; q i represents the power generation flow of the i-th power station; H i，weekly represents the target water level in front of the dam of the i-th power station; h i represents the water level in front of the dam of the i-th power station; fi It represents the function of calculating the final water level in front of the dam of the reservoir of the i-th power station through the initial water level, power generation flow, abandoned water flow and inflow; s i represents the abandoned water flow of the i-th power station; p i represents the output of the i-th power station; s1 represents the abandoned water flow of the first-level power station, p1 represents the output of the first-level power station; q1 represents the power generation flow of the first-level power station; s2 represents the abandoned water flow of the second-level power station; p2 represents the output of the second-level power station; q2 represents the power generation flow of the second-level power station; s3 represents the abandoned water flow of the third-level power station; p3 represents the output of the third-level power station; q3 represents the power generation flow of the third-level power station; ds i represents the adjusted abandoned water flow of the i-th power station; H i represents the lowest water level in front of the dam of the i-th power station; represents the highest water level of the i-th power station; q i represents the minimum power generation flow of the i-th level power station; represents the maximum power generation flow of the i-th power station; s i It represents the minimum water discharge when the gate of the i-th power station is opened; Q i represents the minimum outflow rate of the i-th level power station; represents the maximum outflow of the i-th power station; P daily_plan_pwr Indicates the total planned output of the cascade power station group; nhq i represents the output function of the i-th power station; I1 represents the total inflow of the first-level power station; I j A represents the total inflow of the j-th power station; j represents the natural inflow to the reservoir of the j-th power station; s j-1 represents the abandoned water flow of the j-1th power station; q j-1 V1 represents the power generation flow of the j-1th power station; 0 Represents the initial storage capacity of the first-stage power station; V1 1 End-of-period storage capacity of the first-stage power station; represents the initial storage capacity of the j-th power station; represents the final storage capacity of the j-th power station; s j represents the abandoned water flow of the j-th power station; q j represents the power generation flow of the j-th power station; b i represents the tailwater level of the i-th power station; f i 1 represents the function for calculating the water level in front of the dam based on the storage capacity of the i-th power station, and the function is fitted based on the experimental data of the i-th power station; f i 2It represents the function for calculating the tailwater level based on the power generation flow and the abandoned water flow, and is obtained by fitting the experimental data of power station i.

[0071] Specifically, g3 represents the first optimization goal of dry season scheduling, which means using the deviation value from the weekly water level target as the water level risk measure. The greater the deviation from the weekly water level, the greater the risk. According to the principle of the wooden barrel effect, the risk of multiple power stations depends on the power station with the greatest risk, that is, the power station with the largest deviation. Therefore, first find the power station with the largest deviation, that is, Then minimize it as the scheduling goal, so as to address the shortcomings in risk management.

[0072] g4 represents the second optimization objective of dry season scheduling, representing the benefit index of the dry season, that is, the weighted power generation flow per unit output. The smaller this value, the higher the power generation efficiency. Since water abandonment is generally not allowed during the dry season, a penalty multiplier with a value between [1, 2] is constructed using the tanh function. When there is no water abandonment, there is no penalty and the penalty multiplier is 1. When there is water abandonment, the penalty multiplier is greater than 1, but its value will not exceed 2.

[0073] The current scheduling strategy aims to ensure the successful completion of key tasks within a given future timeframe. However, due to the impact of future uncertainty, a probabilistic assessment of the target task completion is necessary, incorporating future uncertainty factors. Therefore, using future basin water inflow probability forecasts and future load forecasts, and employing Monte Carlo simulation methods, we estimate the impact of current decisions on the achievement of future long-term goals and construct an external utility function. This external utility function quantifies the scheduling strategy as a risk-return characteristic, enabling precise analysis of scheduling simulation information.

[0074] Example 2: Cascade power station scheduling optimization system considering risk and benefit game balance, this method can realize the cascade power station scheduling optimization method considering risk and benefit game balance recorded in Example 1, such as Figure 2 As shown, it includes a scheduling establishment module, a game conversion module, a game solving module, a scheduling simulation module and an evaluation and analysis module.

[0075] The scheduling establishment module is used to establish a multi-task scheduling model for a corresponding time period based on multiple dam statuses, power load plans, future water inflow forecasts, and time factors. The game conversion module is used to establish a corresponding risk-benefit game model based on the scheduling task types in the multi-task scheduling model. It then converts the Pareto efficiency equilibrium strategy problem of the cooperative game between the risk and benefit agents in the risk-benefit game model into a two-objective optimization problem. The game solution module uses the NSGA-III algorithm to solve the Pareto efficiency equilibrium strategy set for the game and uses this Pareto efficiency equilibrium strategy set as the initial scheduling strategy. The scheduling simulation module establishes a corresponding scheduling simulation model based on the scheduling scenario, inputs the initial scheduling strategy into the scheduling simulation model, and collects scheduling simulation information after running it. The evaluation and analysis module evaluates and analyzes the scheduling simulation information based on a pre-constructed external utility function and selects the initial scheduling strategy with the best external utility value from the Pareto non-inferior solutions as the actual scheduling strategy.

[0076] Working principle: The present invention converts the multi-objective optimization problem into a scheduling risk and benefit game problem, realizing the dimensionality reduction of the multi-objective optimization problem; in addition, the risk and benefit decision-making method is introduced into the scheduling, and the multi-objective scheduling problem is converted into a benefit-risk cooperation game problem. Whether it is the improvement of the risk party or the benefit party, it means that the overall benefit has been improved, that is, either under the premise of unchanged benefits, the risk is reduced relative to before; or under the premise of unchanged risks, the benefits are improved compared with all before; or both are improved, so that the scheduling optimization of cascade power stations is more in line with reality, and the accuracy and rationality of cascade power station scheduling are better.

[0077] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. The cascade power station dispatch optimization method considering the risk and benefit game equilibrium is characterized by: The following steps are involved: Establish a multi-task scheduling model within the corresponding time period based on multiple dam statuses, power load plans, future water inflow forecasts, and time factors; According to the scheduling task types in the multi-task scheduling model, a corresponding risk and benefit game model is established, and the Pareto efficiency equilibrium strategy solution problem of the risk and benefit two-subject cooperative game in the risk and benefit game model is converted into a two-objective optimization solution problem. The NSGA-III algorithm is used to solve the Pareto efficiency equilibrium strategy set of the game, and the Pareto efficiency equilibrium strategy set is used as the initial scheduling strategy; Establish a corresponding scheduling simulation model according to the scheduling scenario, input the initial scheduling strategy into the scheduling simulation model and collect the scheduling simulation information after running it; Evaluate and analyze the scheduling simulation information based on the pre-built external utility function, and select the initial scheduling strategy with the best external utility value from the Pareto non-inferior solutions as the actual scheduling strategy; The scheduling scenarios are divided into flood season scheduling and dry season scheduling according to time characteristics; The objective function of the two-objective optimization problem corresponding to the flood season scheduling is specifically: Among them, g1 represents the first optimization goal of flood season scheduling, Indicates the minimum value of multiple calculated values ​​generated by selecting a set of data s1, q1, s2, q2, s3, q3 as input. It represents the maximum value among the multiple minimum values ​​obtained by inputting multiple sets of different data s1, q1, s2, q2, s3, q3; g2 represents the second optimization goal of flood season scheduling, It represents the minimum value of multiple calculated values ​​obtained by selecting multiple sets of different data s1, q1, s2, q2, s3, q3 as input; st represents the constraint condition, C1-C12 represent different constraint conditions; i represents the power station number, the power station number is the power station level; q i represents the power generation flow of the i-th power station; h i represents the water level in front of the dam of the i-th power station; s i represents the abandoned water flow of the i-th power station, p i Indicates the output of the i-th power station; ds i represents the adjusted abandoned water flow of the i-th power station; H i represents the lowest water level in front of the dam of the i-th power station; represents the highest water level of the i-th power station; q i represents the minimum power generation flow of the i-th level power station; represents the maximum power generation flow of the i-th power station; S i It represents the minimum water discharge when the gate of the i-th power station is opened; Q i represents the minimum outflow rate of the i-th level power station; represents the maximum outflow of the i-th power station; P daily_plan_pwr Indicates the total planned output of the cascade power station group; nhq i represents the output function of the i-th power station; I1 represents the total inflow of the first-level power station; I j A represents the total inflow of the j-th power station; j represents the natural inflow to the reservoir of the j-th power station; S j-1 represents the abandoned water flow of the j-1th power station; q j-1 V1 represents the power generation flow of the j-1th power station; 0 Represents the initial storage capacity of the first-stage power station; V1 1 End-of-period storage capacity of the first-stage power station; represents the initial storage capacity of the j-th power station; represents the final storage capacity of the j-th power station; S j represents the abandoned water flow of the j-th power station; q j represents the power generation flow of the j-th power station; b i represents the tailwater level of the i-th power station; f i 1 represents the function for calculating the water level in front of the dam based on the storage capacity of the i-th power station; f i 2 Represents the function for calculating the tailwater level based on the power generation flow and the abandoned water flow.

2. The cascade power station scheduling optimization method considering the risk and benefit game equilibrium according to claim 1 is characterized in that: The contradiction between risks and benefits of flood season scheduling is the contradiction between water storage and flood control.

3. The cascade power station scheduling optimization method considering risk and benefit game equilibrium according to claim 1 is characterized in that: The contradiction between the risks and benefits of dry season scheduling is the contradiction between water consumption rate and water drawdown risk.

4. The cascade power station scheduling optimization method considering risk and benefit game equilibrium according to claim 1 is characterized in that: The risk and benefit game model corresponding to the flood season scheduling is specifically as follows: For flood season benefits, under a given power generation load plan, the smaller the sum of the total outflow of the cascade power station group, the greater the flood season benefits; For flood season risk parties, the safety redundancy of the water level in front of the dam of the cascade hydropower station group is used as a measure. The more the water level is below the safety level, the greater the safety redundancy, the smaller the risk, and the greater the benefit to the risk party.

5. The cascade power station scheduling optimization method considering risk and benefit game equilibrium according to claim 1 is characterized in that: The risk and benefit game model corresponding to the dry season scheduling is specifically as follows: In terms of dry season benefits, the lower the comprehensive water consumption rate of the cascade power station group, the better. The comprehensive water consumption rate is obtained by dividing the total power generation water of all power stations by the total power generation. For the dry season risk aspect, the difference between the planned water level in front of the dam around the power station and the target water level is defined as the measure of the dry season risk. The greater the deviation, the greater the risk.

6. The cascade power station scheduling optimization method considering risk and benefit game equilibrium according to claim 1 is characterized in that: The objective function of the two-objective optimization problem corresponding to the dry season scheduling is specifically: Among them, g3 represents the first optimization goal of dry season scheduling, Indicates the minimum value of multiple calculated values ​​generated by selecting a set of data s1, q1, s2, q2, s3, q3 as input. It represents the maximum value among multiple minimum values ​​obtained by inputting multiple sets of different data s1, q1, s2, q2, s3, q3; g4 represents the second optimization goal of dry season scheduling, It represents the minimum value of multiple calculated values ​​obtained by selecting multiple sets of different data s1, q1, s2, q2, s3, q3 as input; st represents the constraint condition, C1-C12 represent different constraint conditions; i represents the power station number, the power station number is the power station level; q i represents the power generation flow of the i-th power station; H i,weekly represents the target water level in front of the dam of the i-th power station; h i represents the water level in front of the dam of the i-th power station; f i It represents the function of calculating the final water level in front of the dam of the reservoir of the i-th power station through the initial water level, power generation flow, abandoned water flow and inflow; s i represents the abandoned water flow of the i-th power station; p i Indicates the output of the i-th power station; ds i represents the adjusted abandoned water flow of the i-th power station; H i represents the lowest water level in front of the dam of the i-th power station; represents the highest water level of the i-th power station; q i represents the minimum power generation flow of the i-th level power station; represents the maximum power generation flow of the i-th power station; S i It represents the minimum water discharge when the gate of the i-th power station is opened; Q i represents the minimum outflow rate of the i-th level power station; represents the maximum outflow of the i-th power station; P daily_plan_pwr Indicates the total planned output of the cascade power station group; nhq i represents the output function of the i-th power station; I1 represents the total inflow of the first-level power station; I j A represents the total inflow of the j-th power station; j represents the natural inflow to the reservoir of the j-th power station; s j-1 represents the abandoned water flow of the j-1th power station; q j-1 V1 represents the power generation flow of the j-1th power station; 0 Represents the initial storage capacity of the first-stage power station; V1 1 End-of-period storage capacity of the first-stage power station; represents the initial storage capacity of the j-th power station; represents the final storage capacity of the j-th power station; s j represents the abandoned water flow of the j-th power station; q j represents the power generation flow of the j-th power station; b i represents the tailwater level of the i-th power station; f i 1 represents the function for calculating the water level in front of the dam based on the storage capacity of the i-th power station; f i 2 Represents the function for calculating the tailwater level based on the power generation flow and the abandoned water flow.

7. The cascade power station scheduling optimization method considering risk and benefit game equilibrium according to claim 1 is characterized in that: The external utility function is constructed by using the future basin water inflow probability forecast and future load forecast, and adopting the Monte Carlo simulation method to estimate the impact of current decisions on the achievement of future long-term goals, thereby constructing the external utility function.

8. The cascade power station dispatch optimization system considering the risk and benefit game equilibrium is characterized by: include: The scheduling establishment module is used to establish a multi-task scheduling model within the corresponding time period based on multiple dam states, power load plans, future water inflow forecasts and time factors; The game conversion module is used to establish a corresponding risk and benefit game model according to the scheduling task type in the multi-task scheduling model, and convert the Pareto efficiency equilibrium strategy solution problem of the risk and benefit two-subject cooperative game in the risk and benefit game model into a two-objective optimization solution problem; The game solving module is used to solve the Pareto efficiency equilibrium strategy set of the game using the NSGA-III algorithm and use the Pareto efficiency equilibrium strategy set as the initial scheduling strategy; The scheduling simulation module is used to establish a corresponding scheduling simulation model according to the scheduling scenario, input the initial scheduling strategy into the scheduling simulation model and collect scheduling simulation information after running; The evaluation and analysis module is used to evaluate and analyze the scheduling simulation information based on the pre-built external utility function, and select the initial scheduling strategy with the best external utility value from the Pareto non-inferior solutions as the actual scheduling strategy; The scheduling scenarios are divided into flood season scheduling and dry season scheduling according to time characteristics; The objective function of the two-objective optimization problem corresponding to the flood season scheduling is specifically: Among them, g1 represents the first optimization goal of flood season scheduling, Indicates the minimum value of multiple calculated values ​​generated by selecting a set of data s1, q1, s2, q2, s3, q3 as input. It represents the maximum value among the multiple minimum values ​​obtained by inputting multiple sets of different data s1, q1, s2, q2, s3, q3; g2 represents the second optimization goal of flood season scheduling, It represents the minimum value of multiple calculated values ​​obtained by selecting multiple sets of different data s1, q1, s2, q2, s3, q3 as input; st represents the constraint condition, C1-C12 represent different constraint conditions; i represents the power station number, the power station number is the power station level; q i represents the power generation flow of the i-th power station; h i represents the water level in front of the dam of the i-th power station; s i represents the abandoned water flow of the i-th power station, p i Indicates the output of the i-th power station; ds i represents the adjusted abandoned water flow of the i-th power station; H i represents the lowest water level in front of the dam of the i-th power station; represents the highest water level of the i-th power station; q i represents the minimum power generation flow of the i-th level power station; represents the maximum power generation flow of the i-th power station; S i It represents the minimum water discharge when the gate of the i-th power station is opened; Q i represents the minimum outflow rate of the i-th level power station; represents the maximum outflow of the i-th power station; P daily_plan_pwr Indicates the total planned output of the cascade power station group; nhq i represents the output function of the i-th power station; I1 represents the total inflow of the first-level power station; I j A represents the total inflow of the j-th power station; j represents the natural inflow to the reservoir of the j-th power station; S j-1 represents the abandoned water flow of the j-1th power station; q j-1 V1 represents the power generation flow of the j-1th power station; 0 Represents the initial storage capacity of the first-stage power station; V1 1 End-of-period storage capacity of the first-stage power station; represents the initial storage capacity of the j-th power station; represents the final storage capacity of the j-th power station; S j represents the abandoned water flow of the j-th power station; q j represents the power generation flow of the j-th power station; b i represents the tailwater level of the i-th power station; f i 1 represents the function for calculating the water level in front of the dam based on the storage capacity of the i-th power station; f i 2 Represents the function for calculating the tailwater level based on the power generation flow and the abandoned water flow.

9. The cascade power station dispatch optimization system considering risk and benefit game balance according to claim 8 is characterized in that: The risk and benefit game model corresponding to the flood season scheduling is specifically as follows: For flood season benefits, under a given power generation load plan, the smaller the sum of the total outflow of the cascade power station group, the greater the flood season benefits; For flood season risk parties, the safety redundancy of the water level in front of the dam of the cascade hydropower station group is used as a measure. The more the water level is below the safety level, the greater the safety redundancy, the smaller the risk, and the greater the benefit to the risk party.

10. The cascade power station dispatch optimization system considering risk and benefit game balance according to claim 8 is characterized in that: The risk and benefit game model corresponding to the dry season scheduling is specifically as follows: In terms of dry season benefits, the lower the comprehensive water consumption rate of the cascade power station group, the better. The comprehensive water consumption rate is obtained by dividing the total power generation water of all power stations by the total power generation. For the dry season risk aspect, the difference between the planned water level in front of the dam around the power station and the target water level is defined as the measure of the dry season risk. The greater the deviation, the greater the risk.

Images