A multi-objective balanced dispatching method for hydropower stations with minimal negative externalities

Abstract

The present invention discloses a multi-objective balanced dispatching method for a hydropower station with minimal negative externalities, including selecting a dispatching objective function according to actual dispatching needs, and setting constraints and optimization solution parameters to construct a multi-objective optimization dispatching model; solving each non-inferior solution on the Pareto front of the multi-objective optimization dispatching model based on a selected solution algorithm to obtain a non-inferior solution set; using a negative externality calculation method to calculate the negative externalities of each non-inferior solution on the Pareto front, and taking the dispatching scheme corresponding to the non-inferior solution with the minimum negative externality on the non-inferior solution set as the optimal balanced dispatching scheme. The advantages are: effectively balancing the fairness and efficiency among the various objectives, ensuring that while improving the operating efficiency of the hydropower station, other objectives such as ecology can also be fairly considered. Compared with the prior art, this balanced dispatching method significantly improves the "efficiency" and "fairness" of multi-objective dispatching.

Description

Technical Field

[0001] The invention relates to the technical field of multi-objective optimization dispatching of reservoirs, and in particular to a multi-objective balanced dispatching method for a hydropower station with minimum negative externalities. Background Art

[0002] The construction of hydropower stations not only interrupts the natural connectivity of rivers, but also changes the original hydrological situation. As a direct result of human activities, the operation decisions of hydropower stations, while bringing significant benefits such as power generation, water supply and shipping, inevitably have an impact on ecology and other aspects that cannot be ignored. In the process of multi-objective reservoir scheduling, there is an obvious competitive relationship between the various objectives. In the past, the solution to this multi-objective competitive reservoir scheduling problem was mostly to seek non-inferior solutions with the help of the Pareto frontier. However, this method takes the maximum benefit between objectives as the competition principle, and multiple objectives compete fully around this maximum principle. The resulting scheduling plan is the product of pursuing maximum benefits, but ignores the fairness requirements among multiple objectives. In actual multi-objective scheduling, power generation, shipping, water supply, etc. have very clear benefit indicators, while the ecological benefit indicators are relatively vague, with a large elastic space, and the optimal aquatic ecology is to maintain the natural hydrological process, which is naturally in conflict with the scheduling process of hydropower stations. Therefore, the ecological benefits are often easily replaced by other objectives, resulting in scheduling plans that do not fully consider the "fairness" of objectives such as ecology. Summary of the invention

[0003] The purpose of the present invention is to provide a multi-objective balanced scheduling method for a hydropower station with minimal negative externalities, thereby solving the above-mentioned problems existing in the prior art.

[0004] In order to achieve the above object, the technical solution adopted by the present invention is as follows:

[0005] A multi-objective balanced dispatching method for a hydropower station with minimal negative externalities comprises the following steps:

[0006] S1. Construct a multi-objective optimization scheduling model and solve the non-inferior solution on its Pareto frontier:

[0007] Select the objective function of the scheduling according to the actual scheduling needs, and set the constraints and optimization solution parameters to build a multi-objective optimization scheduling model; solve the non-inferior solutions on the Pareto frontier of the multi-objective optimization scheduling model based on the selected solution algorithm to obtain the non-inferior solution set;

[0008] S2. Determine the optimal scheduling plan based on the negative externalities of non-inferior solutions:

[0009] Using the negative externality calculation method, the negative externalities of each non-inferior solution on the Pareto front are calculated, and the scheduling plan corresponding to the non-inferior solution with the smallest negative externality in the non-inferior solution set is taken as the optimal equilibrium scheduling plan;

[0010] The negative externality calculation method calculates the difference ratio of each data point by utilizing the natural flow and the scheduled flow of the reservoir in the same time period, and calculates the average difference ratio of the data points based on the sum of the difference ratios of the data points, and uses the average difference ratio of the data points as an evaluation indicator of the negative externality.

[0011] Preferably, the objective function is at least two of flood control objectives, power generation objectives, shipping objectives, ecological objectives, and water supply objectives.

[0012] Preferably, the constraints include upper and lower limit constraints on reservoir capacity, water level fluctuation constraints, reservoir water balance constraints, reservoir water balance constraints, reservoir discharge capacity constraints, outflow flow constraints, and output constraints.

[0013] Storage capacity upper and lower limit constraints:

[0014]

[0015] in, and are the minimum and maximum water level constraints of the ith reservoir in the tth period during the dispatch period, respectively; To regulate the water level in front of the dam of the i-th reservoir in the t-th period;

[0016] Water level fluctuation constraints:

[0017] |Z i,t -Z i,t+1 |≤ΔZ i , i≥1

[0018] Among them, Z i,t and Z i,t+1 are the water levels of the ith reservoir in the tth period and the t+1th period during the dispatch period; ΔZ i is the upper limit of the daily water level variation of the ith reservoir during the dispatch period;

[0019] Reservoir water balance constraints:

[0020]

[0021] in, and are the storage capacity, inflow and discharge flow of the ith reservoir in the tth period during the dispatch period; is the storage capacity of the ith reservoir in the t-1th period during the dispatch period; Δt is the time difference from water inflow to water discharge;

[0022] Reservoir discharge capacity constraints:

[0023]

[0024] in, is the maximum discharge capacity of the ith reservoir at the dam front water level in the tth period during the dispatch period;

[0025] Outbound traffic constraints:

[0026]

[0027]

[0028] in, is the lower limit of the outflow of the ith reservoir during the dispatch period; is the discharge flow of the i-th reservoir in the t-1th period during the dispatch period, ΔQ i is the upper limit of the daily discharge fluctuation of the i-th reservoir during the dispatch period;

[0029] Output constraints:

[0030]

[0031] in, is the output of the ith reservoir in the tth period during the dispatch period; and They are the minimum and maximum output constraints of the corresponding reservoirs in the same period during the scheduling period.

[0032] Preferably, the optimization solution parameters include step size, population size, number of iterations, and input flow data.

[0033] Preferably, the solution algorithm is a genetic algorithm or a heuristic algorithm.

[0034] Preferably, the negative externality calculation method includes the following contents:

[0035] S31, obtaining the natural flow process of the reservoir and the flow process after the reservoir is dispatched;

[0036] S32, interpolating and replacing missing data, abnormal values ​​and mutation points of the natural flow process and the flow process after reservoir regulation;

[0037] S33. Based on the natural flow sequence and the flow sequence after reservoir scheduling in the same time period and time step, the difference ratio of each data point is calculated, and the average difference ratio of the data points is calculated according to the sum of the difference ratios of the data points, and the average difference ratio of the data points is used as an evaluation indicator of negative externalities.

[0038] Preferably, step S33 specifically includes the following contents:

[0039] S331. Based on the natural flow process after data processing and the flow process after reservoir regulation, obtain the natural flow sequence and the flow sequence after reservoir regulation in the same time period and time step, which are recorded as Q_natural() and Q_current() respectively; each sequence has N data;

[0040] S332. For each data point, the absolute value of the difference between the natural flow and the flow after reservoir regulation is calculated, and then divided by the natural flow to obtain the difference ratio of each data point; the calculation formula is,

[0041]

[0042] Among them, R j is the difference ratio of the jth data point, 1≤j≤N; Q natural,k is the natural traffic data at time k; Q current,k is the flow data after reservoir regulation at time k;

[0043] S333, based on the difference ratios of all data points, the sum of the difference ratios of the data points is calculated, and then divided by the total number of data points to obtain the average difference ratio of the data points; the calculation formula is,

[0044]

[0045] Where R is the average difference ratio of the data points;

[0046] S34. The average difference ratio R of the data points is used as the negative externality evaluation index; R takes a value between 0 and 1, and the smaller the R value is, the smaller the negative externality is; conversely, the greater the negative externality is.

[0047] Preferably, step S31 specifically includes the following contents:

[0048] S311. Obtain the natural flow process without human intervention by constructing a distributed hydrological model or using mathematical statistics methods;

[0049] S312. Obtain the flow process after reservoir scheduling by means of historical data query and real-time data observation.

[0050] Preferably, in step S311, a distributed hydrological model is constructed to obtain a natural flow process without human intervention, which specifically includes the following contents:

[0051] S3111. Construct a distributed hydrological model multi-factor integrated database module to provide a basis for the construction of the hydrological model, wherein the module includes a basic hydrological database and a water conservancy project database of the target basin;

[0052] The historical precipitation and temperature data of the target basin were obtained based on the China precipitation and temperature 0.25° grid data set; the meteorological data of the target basin were obtained based on the NCEP-FNL reanalysis data; the river network and surface elevation data of the target basin were obtained based on the HydroSHEDs data set; the soil texture parameters of the target basin were obtained based on the World Soil Database; the land use data of the target basin were obtained based on the very high resolution radiometer; and a variety of fusion algorithms were combined to scale the obtained data up and down to the same spatial and temporal resolution to build a basic hydrological database covering the target basin;

[0053] Collect basic data of water conservancy projects in the target basin, including basic engineering characteristic data of reservoirs, dispatching diagrams, and dispatching regulations, and construct a water conservancy project database for the target basin based on the data;

[0054] S3112. Input the data from the basic hydrological database and water conservancy project database of the target basin into the distributed hydrological model to construct a distributed hydrological model of the target area; calibrate the runoff parameters of the distributed hydrological model of the target area based on the measured runoff of the hydrological station, and verify the flow data obtained by simulating the distributed hydrological model of the target area, requiring that the flow certainty coefficient of the simulated flow at the daily scale is greater than the preset threshold; finally, obtain the natural flow process without human intervention based on the calibrated distributed hydrological model of the target area.

[0055] Preferably, step S32 specifically includes the following contents:

[0056] S321. For the missing data in the flow process, the linear interpolation method is used to process it to ensure the continuity of the runoff data;

[0057] S322. For outliers and mutation points in flow process data, linear interpolation method is used to replace them to improve data rationality.

[0058] The beneficial effects of the present invention are: 1. Based on the traditional Pareto front multi-objective scheduling method, the method of the present invention introduces the index of negative externalities, and achieves the balance or coordination of the scheduling scheme in terms of "efficiency" and "fairness" by seeking the scheme with the minimum negative externalities on the Pareto front. 2. The method of the present invention solves the problem that the existing technology only pursues the maximization of multi-objective benefits and ignores the need for fairness among multiple objectives when dealing with multi-objective scheduling problems (such as power generation, shipping, water supply and ecology). By introducing the concept of negative externalities, the scheduling scheme with the minimum negative externalities on the Pareto front is sought to achieve a balance among the multiple objectives of the scheduling scheme. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 is a flow chart of a scheduling method in an embodiment of the present invention;

[0060] Figure 2 is a flow chart of a method for calculating negative externalities in an embodiment of the present invention;

[0061] Figure 3 Schematic diagram of the Pareto front in an embodiment of the present invention. DETAILED DESCRIPTION

[0062] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation methods described herein are only used to explain the present invention and are not used to limit the present invention.

[0063] Embodiment 1

[0064] like Figure 1 As shown, in this embodiment, a multi-objective balanced scheduling method for a hydropower station with minimal negative externalities is provided. First, a multi-objective optimization scheduling model is constructed, including objective functions, constraints and parameter settings. Then, a certain genetic algorithm is selected to solve and obtain the Pareto front solution set. By introducing the negative externality index, the negative externalities of all scheduling schemes in the non-inferior solution set are calculated. Finally, the scheduling scheme with the minimum negative externalities is taken as the optimal scheduling scheme. This method overcomes the limitations existing in the prior art, and searches for the scheduling scheme with the minimum total negative externalities on the Pareto frontier, effectively balancing the fairness and efficiency among the various objectives, ensuring that while improving the operating efficiency of the hydropower station, other objectives such as ecology can also be fairly considered. Compared with the prior art, this balanced scheduling method significantly improves the "efficiency" and "fairness" of multi-objective scheduling. The method of the present invention comprises the following steps,

[0065] 1. Construct a multi-objective optimization scheduling model and solve the non-inferior solution on its Pareto frontier:

[0066] According to the actual scheduling needs, select the scheduling objective function, set the constraints and optimization parameters, and build a multi-objective optimization scheduling model; solve the non-inferior solutions on the Pareto frontier of the multi-objective optimization scheduling model based on the selected solution algorithm to obtain the non-inferior solution set. The specific steps include the following:

[0067] 1.1. Selection of objective function: Determine the objective function of scheduling based on actual scheduling needs, including flood control, power generation, shipping, ecology, water supply and other objectives.

[0068] 1.2. Setting of constraints: including upper and lower limit constraints of reservoir capacity (water level), water level fluctuation constraints, reservoir water balance constraints, reservoir water balance constraints, reservoir discharge capacity constraints, outflow flow constraints, output constraints, etc.

[0069] 1.3. Setting of optimization solution parameters: including step size, population size, number of iterations, and input flow data.

[0070] 1.4. Select the solution algorithm: You can choose genetic algorithm or other heuristic algorithm.

[0071] 1.5. Solve the non-inferior solution set: Based on the solution algorithm and related parameter settings, the multi-objective optimization scheduling model is solved and calculated to obtain the Pareto frontier and the non-inferior solution set.

[0072] 2. Determine the optimal scheduling plan based on the negative externalities of non-inferior solutions:

[0073] Using the negative externality calculation method, the negative externalities of each non-inferior solution on the Pareto front are calculated, and the scheduling scheme corresponding to the non-inferior solution with the smallest negative externality in the non-inferior solution set is taken as the optimal equilibrium scheduling scheme. The specific steps include:

[0074] 2.1. Calculate the negative externalities of all solutions on the non-inferior solution set: Through the negative externality calculation method, the total negative externalities of each non-inferior solution on the Pareto frontier are obtained.

[0075] 2.2. Select the optimal scheduling plan: Find the scheduling plan corresponding to the solution with the minimum total negative externality in the non-inferior solution set as the optimal equilibrium scheduling plan.

[0076] In this embodiment, the negative externality calculation method calculates the difference ratio of each data point by using the natural flow and the scheduled flow of the reservoir in the same time period, and calculates the average difference ratio of the data point according to the sum of the difference ratios of the data points, and uses the average difference ratio of the data points as the evaluation index of the negative externality. Figure 2 As shown in Figure 2, the negative externality calculation method specifically includes the following steps:

[0077] (1) Flow data acquisition: Obtain the natural flow process of the reservoir and the flow process after the reservoir is dispatched. Specifically,

[0078] (1.1) Obtaining natural flow processes: Obtaining natural flow processes without human intervention by constructing distributed hydrological models or using mathematical and statistical methods.

[0079] In this embodiment, the natural flow process without human intervention is obtained by constructing a distributed hydrological model.

[0080] (1.1.1) Construct a distributed hydrological model multi-factor comprehensive database module, including basic hydrological database and water conservancy project database, including:

[0081] a. Basic hydrological database: The historical precipitation and temperature data of the basin are obtained based on the China precipitation and temperature 0.25° grid data set; the meteorological data of the basin, such as air pressure, radiation, specific humidity, and wind speed, are obtained based on the NCEP-FNL reanalysis data; the river network and surface elevation (DEM) data of the basin are obtained based on the HydroSHEDs data set; the soil texture parameters of the basin are obtained based on the World Soil Database (HWSD); the land use data of the basin are obtained based on the Very High Resolution Radiometer (AVHRR); and a variety of fusion algorithms are combined to scale the above data up and down to the same spatiotemporal resolution to construct a basic hydrological database covering a certain basin.

[0082] b. Water conservancy project database: Collect the basic data of regional water conservancy projects to form a water conservancy project database. The basic data includes the basic engineering characteristic data of the reservoir (characteristic storage capacity, characteristic water level, water level storage capacity curve, discharge curve, etc.), dispatching diagram, dispatching regulations, and the water conservancy project database of the basin is constructed based on this, providing a basis for the construction of the hydrological model.

[0083] (1.1.2) Construct a distributed hydrological model of the basin, including:

[0084] a. Input data such as DEM, meteorological drivers, vegetation cover and soil texture into the distributed hydrological model to construct a distributed hydrological model for the area.

[0085] b. Calibrate the runoff parameters of the distributed hydrological model based on the measured runoff at the hydrological station.

[0086] c. Verify the flow data obtained by distributed hydrological model simulation. The flow certainty coefficient of daily scale simulation is required to be greater than 0.8.

[0087] d. Finally, the distributed hydrological model after calibration is used to obtain the natural flow process of the basin without human intervention.

[0088] In this embodiment, real-time monitoring and acquisition and historical data are integrated in the target area and transmitted to the database to achieve rolling update of the database.

[0089] (1.2) Obtain the flow process after reservoir scheduling: Obtain the flow process after reservoir scheduling through historical data query and real-time data observation.

[0090] (2) Flow data processing: interpolation and replacement of missing data, outliers and mutation points in natural flow processes and flow processes after reservoir regulation. Specifically,

[0091] (2.1) Interpolation processing: For the missing data in the flow process, linear interpolation method is used to ensure the continuity of runoff data.

[0092] (2.2) Replacement processing: For abnormal values ​​(such as negative values) and mutation points in the flow process data, linear interpolation method is used to replace them to improve data rationality.

[0093] (3) Determination of negative externalities: Based on the natural flow sequence and the flow sequence after reservoir regulation in the same time period and time step, the difference ratio of each data point is calculated, and the average difference ratio of the data points is calculated based on the sum of the difference ratios of the data points. The average difference ratio of the data points is used as the evaluation indicator of negative externalities. Specifically,

[0094] (3.1) Based on the natural flow process after data processing and the flow process after reservoir regulation, obtain the natural flow sequence and the flow sequence after reservoir regulation with the same time period and time step, which are recorded as Q_natural() and Q_current() respectively; there are N data in each sequence;

[0095] (3.2) For each data point, the absolute value of the difference between the natural flow and the flow after reservoir regulation is calculated, and then divided by the natural flow to obtain the difference ratio of each data point; the calculation formula is,

[0096]

[0097] Among them, R j is the difference ratio of the jth data point, 1≤j≤N; Q natural,k is the natural traffic data at time k; Q current,k is the flow data after reservoir regulation at time k;

[0098] (3.3) Based on the difference ratios of all data points, the sum of the difference ratios of the data points is calculated, and then divided by the total number of data points to obtain the average difference ratio of the data points; the calculation formula is,

[0099]

[0100] Where R is the average difference ratio of the data points;

[0101] (3.4) The average difference ratio R of the data points is used as the negative externality evaluation index; R ranges from 0 to 1. The smaller the R value, the smaller the negative externality; conversely, the larger the negative externality.

[0102] Embodiment 2

[0103] In this embodiment, according to the steps described in the method of the present invention, a cascade reservoir is selected as a case to determine the optimal scheduling plan.

[0104] 1. Construct a multi-objective optimization scheduling model and solve the non-inferior solution on its Pareto frontier:

[0105] 1.1. Selection of objective function: According to the actual needs of the reservoir, two scheduling objectives are selected: maximum power generation and minimum ecological water supply shortage. The objective functions are as follows:

[0106] a. Power generation target: expressed as the maximum total power generation of the cascade reservoirs during the calculation period.

[0107]

[0108] Wherein, T is the total number of calculation periods; n is the number of reservoirs in the cascade group; Ni(t) is the output of the i-th reservoir in the t-th period, Ki is the output coefficient, Qi(t) is the power generation flow, Hi(t) is the average power generation head; Δt is the length of the calculation period.

[0109] b. Minimum ecological water shortage:

[0110]

[0111] Among them, Q eco,i is the minimum ecological flow of the ith reservoir.

[0112] 1.2. Setting of constraints: Constraints include upper and lower limit constraints of reservoir capacity, water level fluctuation constraints, reservoir water balance constraints, reservoir water balance constraints, reservoir discharge capacity constraints, outflow flow constraints, and output constraints.

[0113] a. Storage capacity upper and lower limit constraints:

[0114]

[0115] in, and are the minimum and maximum water level constraints of the ith reservoir in the tth period during the dispatch period, respectively; To regulate the water level in front of the dam of the i-th reservoir in the t-th period;

[0116] b. Water level fluctuation constraint:

[0117] |Z i,t -Z i,t+1 |≤ΔZ i , i≥1

[0118] Among them, Z i,t and Z i,t+1 are the water levels of the ith reservoir in the tth period and the t+1th period during the dispatch period; ΔZ i is the upper limit of the daily water level variation of the ith reservoir during the dispatch period;

[0119] c. Reservoir water balance constraints:

[0120]

[0121] in, and are the storage capacity, inflow and discharge flow of the ith reservoir in the tth period during the dispatch period; is the storage capacity of the ith reservoir in the t-1th period during the scheduling period; Δt is the length of the calculation period;

[0122] d. Reservoir discharge capacity constraints:

[0123]

[0124] in, is the maximum discharge capacity of the ith reservoir at the dam front water level in the tth period during the dispatch period;

[0125] e. Outbound flow constraints:

[0126]

[0127]

[0128] in, is the lower limit of the outflow of the ith reservoir during the dispatch period; is the discharge flow of the i-th reservoir in the t-1th period during the dispatch period, ΔQ i is the upper limit of the daily discharge fluctuation of the i-th reservoir during the dispatch period;

[0129] f. Output constraints:

[0130]

[0131] in, is the output of the ith reservoir in the tth period during the dispatch period; and They are the minimum and maximum output constraints of the corresponding reservoirs in the same period during the scheduling period.

[0132] 1.3、Optimize the setting of solution parameters:

[0133] a. The step length is 1 day.

[0134] b. The scheduling period is from April 1st to July 10th.

[0135] c. The population size is 3000

[0136] d. The number of iterations is 10,000.

[0137] e. Input data is the inflow of Longtou Reservoir and the interval flow between reservoirs (m 3 / s).

[0138] 1.4. Select the solution algorithm: The solution algorithm adopts NSGA3 algorithm.

[0139] 1.5. Solve the non-inferior solution set: Use the NSGA3 algorithm and related parameter settings to solve the multi-objective optimization scheduling model and obtain the Pareto frontier as follows: Figure 3 As shown, there are 3000 solutions in total.

[0140] 2. Determine the optimal scheduling plan based on the negative externalities of non-inferior solutions:

[0141] 2.1. Calculate the negative externalities of all solutions on the non-inferior solution set. Using the existing negative externality calculation method, we can solve the negative externalities of each non-inferior solution on the Pareto frontier.

[0142] 2.2. Select the optimal scheduling plan. Find the scheduling plan corresponding to the solution with the minimum negative externality in the non-inferior solution set as the optimal scheduling plan A.

[0143] By adopting the above technical solution disclosed in the present invention, the following beneficial effects are obtained:

[0144] The present invention provides a hydropower multi-objective balanced scheduling method with minimal negative externalities. The method of the present invention introduces an indicator of negative externalities on the basis of the traditional multi-objective scheduling method of the Pareto front, and achieves a balance or coordination between "efficiency" and "fairness" in the scheduling scheme by seeking a scheme with the minimum negative externalities on the Pareto front. The method of the present invention solves the problem that the prior art only pursues the maximization of multi-objective benefits and ignores the need for fairness among multiple objectives when dealing with multi-objective scheduling problems (such as power generation, shipping, water supply, and ecology). By introducing the concept of negative externalities, a scheduling scheme with the minimum negative externalities on the Pareto front is sought to achieve a balance among multiple objectives of the scheduling scheme.

[0145] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be considered as the scope of protection of the present invention.

Claims

1. A multi-objective balanced dispatching method for a hydropower station with minimal negative externalities, characterized by: The following steps are included: S1. Construct a multi-objective optimization scheduling model and solve the non-inferior solution on its Pareto frontier: Select the objective function of the scheduling according to the actual scheduling needs, and set the constraints and optimization solution parameters to build a multi-objective optimization scheduling model; solve the non-inferior solutions on the Pareto frontier of the multi-objective optimization scheduling model based on the selected solution algorithm to obtain the non-inferior solution set; The objective function is at least two of the following: flood control objective, power generation objective, shipping objective, ecological objective, and water supply objective; specifically, the objective function is to maximize power generation and minimize ecological water supply deficit. a. Power generation target: expressed as the maximum total power generation of cascade reservoirs during the calculation period. Wherein, T is the total number of calculation periods; n is the number of reservoirs in the cascade group; Ni(t) is the output of the i-th reservoir in the t-th period, Ki is the output coefficient, Qi(t) is the power generation flow, Hi(t) is the average power generation head; Δt is the length of the calculation period; b. Minimum ecological water shortage: Among them, Q eco,i is the minimum ecological flow of the i-th reservoir; Q out,i (t) is the discharge flow of the i-th reservoir in the t-th period; S2. Determine the optimal scheduling plan based on the negative externalities of non-inferior solutions: Using the negative externality calculation method, the negative externalities of each non-inferior solution on the Pareto front are calculated, and the scheduling plan corresponding to the non-inferior solution with the smallest negative externality in the non-inferior solution set is taken as the optimal equilibrium scheduling plan; The negative externality calculation method calculates the difference ratio of each data point by using the natural flow and the dispatched flow of the reservoir in the same time period, and calculates the average difference ratio of the data point according to the sum of the difference ratios of the data points, and uses the average difference ratio of the data points as an evaluation index of the negative externality; The negative externality calculation method includes the following contents: S31, obtaining the natural flow process of the reservoir and the flow process after the reservoir is dispatched; S32, interpolation and replacement of missing data, abnormal values ​​and mutation points in the natural flow process and the flow process after reservoir regulation; Step S32 specifically includes the following contents: S321. For the missing data in the flow process, the linear interpolation method is used to process it to ensure the continuity of the runoff data; S322. For abnormal values ​​and mutation points in the flow process data, linear interpolation method is used to replace them to improve data rationality; S33, based on the natural flow sequence and the flow sequence after reservoir regulation in the same time period and time step, calculate the difference ratio of each data point, and calculate the average difference ratio of the data point according to the sum of the difference ratios of the data points, and use the average difference ratio of the data points as an evaluation indicator of negative externalities; Step S33 specifically includes the following contents: S331. Based on the natural flow process after data processing and the flow process after reservoir regulation, obtain the natural flow sequence and the flow sequence after reservoir regulation in the same time period and time step, which are recorded as Q_natural() and Q_current() respectively; each sequence has N data; S332. For each data point, the absolute value of the difference between the natural flow and the flow after reservoir regulation is calculated, and then divided by the natural flow to obtain the difference ratio of each data point; the calculation formula is, Among them, R j is the difference ratio of the jth data point, 1≤j≤N; Q natural,k is the natural flow data at time k; Q current,k is the flow data after reservoir regulation at time k; S333, based on the difference ratios of all data points, the sum of the difference ratios of the data points is calculated, and then divided by the total number of data points to obtain the average difference ratio of the data points; the calculation formula is, Where R is the average difference ratio of the data points; S34. The average difference ratio R of the data points is used as the negative externality evaluation index; R takes a value between 0 and 1, and the smaller the R value is, the smaller the negative externality is; conversely, the greater the negative externality is.

2. The multi-objective balanced dispatching method for a hydropower station with minimal negative externalities according to claim 1 is characterized by: The constraints include upper and lower limits of reservoir capacity, water level fluctuation constraints, reservoir water balance constraints, reservoir discharge capacity constraints, outflow constraints, and output constraints; Storage capacity upper and lower limit constraints: in, and are the minimum and maximum water level constraints of the ith reservoir in the tth period during the dispatch period, respectively; To regulate the water level in front of the dam of the i-th reservoir in the t-th period; Water level fluctuation constraints: |From i,t -WITH i,t+1 |≤ΔZ i ,i≥1 Among them, Z i,t and Z i,t+1 are the water levels of the ith reservoir in the tth period and the t+1th period during the dispatch period; ΔZ i is the upper limit of the daily water level variation of the ith reservoir during the dispatch period; Reservoir water balance constraints: in, and are the storage capacity, inflow and discharge flow of the ith reservoir in the tth period during the dispatch period; is the storage capacity of the ith reservoir in the t-1th period during the dispatch period; Δt is the time difference from water inflow to water discharge; Reservoir discharge capacity constraints: in, is the maximum discharge capacity of the ith reservoir at the dam front water level in the tth period during the dispatch period; Outbound traffic constraints: in, is the lower limit of the outflow of the ith reservoir during the dispatch period; is the discharge flow of the i-th reservoir in the t-1th period during the dispatch period, ΔQ i is the upper limit of the daily discharge fluctuation of the i-th reservoir during the dispatch period; Output constraints: in, is the output of the ith reservoir in the tth period during the dispatch period; and They are the minimum and maximum output constraints of the corresponding reservoirs in the same period during the scheduling period.

3. The multi-objective balanced dispatching method for a hydropower station with minimal negative externalities according to claim 1 is characterized by: Step S31 specifically includes the following contents: S311. Obtain the natural flow process without human intervention by constructing a distributed hydrological model or using mathematical statistics methods; S312. Obtain the flow process after reservoir scheduling by means of historical data query and real-time data observation.

4. The multi-objective balanced dispatching method for a hydropower station with minimal negative externalities according to claim 3 is characterized by: In step S311, a distributed hydrological model is constructed to obtain the natural flow process without human intervention, which specifically includes the following contents: S3111. Construct a distributed hydrological model multi-factor integrated database module to provide a basis for the construction of the hydrological model, wherein the module includes a basic hydrological database and a water conservancy project database of the target basin; The historical precipitation and temperature data of the target basin were obtained based on the China precipitation and temperature 0.25° grid data set; the meteorological data of the target basin were obtained based on the NCEP-FNL reanalysis data; the river network and surface elevation data of the target basin were obtained based on the HydroSHEDs data set; the soil texture parameters of the target basin were obtained based on the World Soil Database; and the land use data of the target basin were obtained based on the very high resolution radiometer. Combine multiple fusion algorithms to upscale and downscale the acquired data to the same spatial and temporal resolution, and build a basic hydrological database covering the target basin; Collect basic data of water conservancy projects in the target basin, including basic engineering characteristic data of reservoirs, dispatching diagrams, and dispatching regulations, and construct a water conservancy project database for the target basin based on the data; S3112. Input the data from the basic hydrological database and water conservancy project database of the target basin into the distributed hydrological model to construct a distributed hydrological model of the target area; calibrate the runoff parameters of the distributed hydrological model of the target area based on the measured runoff of the hydrological station, and verify the flow data obtained by simulating the distributed hydrological model of the target area, requiring that the flow certainty coefficient of the simulated flow at the daily scale is greater than the preset threshold; finally, obtain the natural flow process without human intervention based on the calibrated distributed hydrological model of the target area.

5. The multi-objective balanced dispatching method for a hydropower station with minimal negative externalities according to claim 1 is characterized by: The optimization solution parameters include step size, population size, number of iterations, and input flow data.

6. The multi-objective balanced dispatching method for a hydropower station with minimal negative externalities according to claim 5 is characterized by: The solution algorithm is a genetic algorithm or a heuristic algorithm.

Images