A method, device, equipment, medium and product for optimizing scheduling of water and electricity

Abstract

The present application relates to the technical field of water and electricity dispatching, and discloses a water and electricity optimization dispatching method, device, equipment, medium and product, comprising: calculating and obtaining a water and sediment ecological evaluation index dataset; combining an operation state dataset, the water and sediment ecological evaluation index dataset, and prediction data to determine a dispatching strategy population; calculating water and sediment ecological evaluation index simulation datasets and operation state simulation datasets of different individual dispatching strategies; calculating the power generation income value, the sediment risk value, the ecological loss value and the constraint violation amount of each individual according to the water and sediment ecological evaluation index simulation datasets and the operation state simulation datasets, and finally determining an optimal dispatching strategy set to dispatch the cascade reservoirs, wherein the present application overcomes the technical problems of the traditional dispatching, such as the mutual separation of the water and sediment process and the ecological evaluation, and the long time-consuming calculation, by means of the fusion of the sediment ecological coupling index and the optimization algorithm collaborative calculation, and realizes the collaborative optimization of the cascade hydropower generation benefit, the sediment balance and the ecological protection.

Description

Technical Field

[0001] This invention relates to the field of hydropower dispatching technology, specifically to a method, apparatus, equipment, medium, and product for optimized hydropower dispatching. Background Technology

[0002] Traditional cascade reservoir operation primarily focuses on power generation efficiency and flood control, paying insufficient attention to sediment transport and riverbed scouring and deposition, and lacking a systematic guarantee for ecological flow. Some studies have begun to introduce ecological flow or sediment concentration as additional constraints into the operation model, but methods that study sediment or ecology in isolation struggle to capture the dynamic coupling relationship between the two. Furthermore, existing optimization methods face bottlenecks when dealing with high-dimensional, multi-constraint, and multi-objective cascade operation problems, including difficulties in generating initial feasible solutions, insufficient population diversity, and long computational times, making it difficult to balance engineering feasibility, ecological protection, and computational efficiency. Summary of the Invention

[0003] This invention provides a method, apparatus, equipment, medium, and product for optimized hydropower scheduling, in order to solve the problem that water and sediment processes and ecological responses are disconnected in traditional cascade reservoir scheduling, making it difficult to achieve multi-objective synergistic optimization.

[0004] In a first aspect, the present invention provides a method for optimized hydropower scheduling, comprising: The system obtains the operational status dataset representing the current operational status of the cascade reservoirs, as well as the inflow forecast data and sediment concentration forecast data for multiple future times. Based on the operational status dataset, it calculates the water and sediment ecological evaluation index dataset representing the water and sediment ecological status. By combining the operational status dataset, the water and sediment ecological evaluation index dataset, and the water inflow forecast data and sediment concentration forecast data for multiple future times, the scheduling strategy population is determined. The scheduling strategy population includes multiple individuals, and each individual corresponds to a scheduling strategy. The simulation datasets of water and sediment ecological evaluation indicators and the simulation datasets of operating status are calculated separately for each future time after scheduling the current operating status using different individual corresponding scheduling strategies. Based on the simulation dataset of water and sediment ecological evaluation indicators and the simulation dataset of operational status, the power generation revenue value, sediment risk value and ecological loss value corresponding to each individual are calculated respectively. The constraint violation amount is determined based on the various types of operational state simulation data and the corresponding constraint boundaries in the operational state simulation dataset. The optimal scheduling strategy set is determined from the current population based on the constraint violation amount and the power generation revenue value, sediment risk value, and ecological loss value corresponding to each individual. If the optimal scheduling strategy set meets the preset conditions, the cascade reservoirs are scheduled based on the optimal scheduling strategy set.

[0005] By constructing a dataset of water and sediment ecological evaluation indicators to characterize the water and sediment ecological state and participating in subsequent calculations, the sediment transport process and the ecological response process are integrated into the scope of scheduling decisions. This overcomes the technical defects of traditional scheduling, which separates water and sediment processes from ecological evaluation and makes it difficult to capture their dynamic coupling relationship. It achieves synergistic optimization of water and sediment processes and ecological responses. Furthermore, by combining sediment transport deviation, scouring and deposition risk, ecological suitability, and ecological sediment coordination with constraint violation quantities, the power generation revenue value, sediment risk value, and ecological loss value are calculated together and optimized. This allows the final output scheduling strategy set to take into account power generation benefits, sediment balance, and ecological protection. It provides an intelligent scheduling scheme that takes into account the synergy of multiple water, sediment, and ecological objectives for cascade reservoir scheduling. In addition, by introducing a scouring and deposition ecological response function to replace the complete physical coupling model for rapid evaluation of candidate schemes, the technical obstacle of excessive computation time and inability to be applied online due to calling the complete water and sediment ecological coupling model is overcome. This allows the algorithm to complete large-scale iterative optimization within a limited time.

[0006] In one optional implementation, the scheduling strategy population is determined by combining the operational status dataset, the water and sediment ecological evaluation index dataset, and the inflow and sediment concentration forecast data for multiple future time periods, including: The operational status dataset, the water and sediment ecological evaluation index dataset, and the water inflow forecast data and sediment concentration forecast data for multiple future times are input into the preset scheduling strategy model for calculation to obtain the initial scheduling strategy. The initial scheduling strategy population is generated by using the initial scheduling strategy as the population center individual, and the population is optimized to obtain the optimized scheduling strategy population.

[0007] By inputting operational status data, water and sediment ecological evaluation index data, and future water and sediment inflow forecast data into a pre-set scheduling strategy model, the model generates an initial scheduling strategy based on a full understanding of the current water and sediment ecological state and future water and sediment inflow trends. Then, an initial population is generated with this strategy as the central individual and optimization calculations are performed. This overcomes the problems of low feasible solution ratio and poor search efficiency caused by the random generation of the initial population in traditional scheduling optimization. It makes the individuals in the initial population physically feasible and objectively superior, thereby improving optimization efficiency and the quality of the final scheduling strategy set.

[0008] In one optional implementation, the simulation datasets of water and sediment ecological evaluation indicators and operational status simulation datasets for future time periods are calculated separately after scheduling the current operational status using different individual-specific scheduling strategies, including: The current operational status dataset and the scheduling strategies corresponding to different individuals are input into a preset scouring and sedimentation ecological response function for calculation, thereby obtaining a simulated dataset of water and sediment ecological evaluation indicators and an operational status simulation dataset for future time periods.

[0009] By inputting the current operating status and different candidate scheduling strategies into a preset scouring and sedimentation ecological response function, the water and sediment ecological evaluation indicators and operating status simulation data at future times after the implementation of each strategy can be obtained quickly in a short period of time. This overcomes the technical bottleneck of traditional methods, which require too much time to call the complete water and sediment ecological coupling model each time and cannot support large-scale iterative optimization. It provides an efficient computational foundation for the collaborative optimization of reinforcement learning of strategy models and strategy optimization algorithms.

[0010] In one optional implementation, the method for constructing the preset scouring and silting ecological response function includes: Acquire multiple training sample data, including historical running status data and historical scheduling strategy data; Training sample data is input into the water-sediment-ecology coupling model for simulation, and the simulation results of the operational state corresponding to each training sample are obtained. Based on the simulation results of the operational state, the simulation results of the water-sediment-ecology evaluation index corresponding to each training sample are calculated. The water-sediment-ecology coupling model includes a hydrodynamic calculation sub-model, a sediment transport calculation sub-model, a riverbed scour and deposition evolution sub-model, and an ecological response evaluation sub-model. The hydrodynamic calculation sub-model is used to simulate water level, flow velocity, water depth, and flow rate according to the scheduling strategy and boundary conditions. The sediment transport calculation sub-model is used to simulate sediment concentration and sediment transport rate according to the hydrodynamic simulation results. The riverbed scour and deposition evolution sub-model is used to simulate scour and deposition volume and riverbed elevation changes according to the sediment transport simulation results. The ecological response evaluation sub-model is used to calculate the ecological suitability index based on the hydrodynamic simulation results, sediment transport simulation results, and riverbed scour and deposition evolution simulation results. Using training sample data as input, and the simulation results of the running state dataset and water and sediment ecological evaluation index corresponding to each training sample as output, the scouring and sedimentation ecological response function to be trained is trained to obtain the trained scouring and sedimentation ecological response function.

[0011] By acquiring training sample data containing historical operating states and historical scheduling strategies, and inputting it into a complete water-sediment-ecological coupling model that includes four sub-models—hydrodynamics, sediment transport, riverbed scouring and deposition, and ecological response—a high-quality input and output sample library covering various operating conditions is generated. Based on this sample library, the scouring and deposition ecological response function is trained. This allows the trained response function to quickly replace the complete physical model, which takes several hours, during the online scheduling phase. This preserves the physical mechanism of water-sediment-ecological coupling while improving computational efficiency.

[0012] In one optional implementation, the operating status simulation dataset also includes unit output data, unit start-up and shutdown cost data, and water wastage loss data. The water and sediment ecological evaluation index simulation dataset includes simulated sediment transport intensity index, simulated scouring and deposition risk index, simulated ecological suitability index, and simulated ecological sediment coupling index. Based on the water and sediment ecological evaluation index simulation dataset and the operating status simulation dataset, the power generation revenue value, sediment risk value, and ecological loss value corresponding to each individual are calculated, including: Obtain electricity cost data and target sediment transport intensity index at various future time points; The power generation output value is determined based on the unit output data and electricity cost data at each time point, and the power generation revenue value at each time point is determined by combining the unit start-up and shutdown cost data and water wastage loss data. The simulated sediment transport deviation is determined based on the simulated sediment transport intensity index and the target sediment transport intensity index at each time point, and the sediment risk value is determined based on the simulated sediment transport deviation and the simulated erosion and deposition risk index. The ecological loss value is determined based on the simulated ecological suitability index and the simulated ecological sediment coupling index.

[0013] In one optional implementation, an optimal scheduling strategy set is determined from the current population based on the constraint violation amount and the power generation revenue value, sediment risk value, and ecological loss value corresponding to each individual, including: The penalty amounts for power generation revenue, sediment risk, and ecological loss are determined by multiplying the penalty coefficients corresponding to power generation revenue, sediment risk, and ecological loss values ​​by the amount of constraint violation. The power generation revenue value, sediment risk value, and ecological loss value, after penalty correction, are determined based on the power generation revenue value, sediment risk value, and ecological loss value, as well as the corresponding penalty amount. The power generation revenue, sediment risk, and ecological loss values ​​of each individual in the current population are non-dominated and sorted. The first preset number of individuals after sorting are taken as the optimal scheduling strategy set for the current population.

[0014] By multiplying the constraint violation by the penalty coefficient corresponding to each objective and then summing them to the power generation revenue value, sediment risk value, and ecological loss value, individuals that violate the constraints are penalized in all three objectives. As a result, they are naturally at a disadvantage in the non-dominated ranking and are prioritized for elimination, without the need to set a separate feasibility priority rule. At the same time, by performing multi-objective trade-offs on the three objective values ​​after penalty correction through non-dominated ranking, the selected optimal scheduling strategy set not only meets the basic constraints of engineering safety and ecological protection, but also achieves an objective balance among the three objectives of power generation, sediment, and ecology without the need for preset weights, thus ensuring the engineering feasibility and multi-objective optimality of the final output strategy.

[0015] In a second aspect, the present invention provides a hydropower optimization scheduling device, comprising: The data acquisition module is used to acquire the operational status dataset representing the current operational status of the cascade reservoirs, as well as the inflow forecast data and sediment concentration forecast data for multiple future times, and to calculate the water and sediment ecological evaluation index dataset representing the water and sediment ecological status based on the operational status dataset. The strategy calculation module is used to combine the operational status dataset, the water and sediment ecological evaluation index dataset, and the water inflow forecast data and sediment concentration forecast data for multiple future times to determine the scheduling strategy population. The scheduling strategy population includes multiple individuals, and each individual corresponds to a scheduling strategy. The simulation calculation module is used to calculate the simulated datasets of water and sediment ecological evaluation indicators and the simulated datasets of operating status at future times after scheduling the current operating status using different individual corresponding scheduling strategies. The target calculation module is used to calculate the power generation revenue, sediment risk, and ecological loss value for each individual based on the simulation dataset of water and sediment ecological evaluation indicators and the simulation dataset of operational status. The constraint violation calculation module is used to determine the constraint violation amount based on various types of simulation data of running states and the corresponding constraint boundaries in the running state simulation dataset. The strategy ranking module is used to determine the optimal scheduling strategy set from the current population based on the constraint violation amount and the power generation revenue value, sediment risk value and ecological loss value corresponding to each individual. The scheduling execution module is used to schedule cascade reservoirs based on the optimal scheduling strategy set if the optimal scheduling strategy set meets preset conditions.

[0016] Thirdly, the present invention provides an electronic device, comprising: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to perform a hydropower optimization scheduling method of the first aspect or any corresponding embodiment described above.

[0017] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to execute a hydropower optimization scheduling method according to the first aspect or any corresponding embodiment described above.

[0018] Fifthly, the present invention provides a computer program product, including computer instructions, which are used to cause a computer to execute a hydropower optimization scheduling method according to the first aspect above or any corresponding embodiment thereof. Attached Figure Description

[0019] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of an application scenario according to an embodiment of the present invention; Figure 2 This is a flowchart illustrating a hydropower optimization scheduling method according to an embodiment of the present invention; Figure 3 This is a structural block diagram of a hydropower optimization scheduling device according to an embodiment of the present invention; Figure 4 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] It is understood that before using the technical solutions disclosed in the various embodiments of the present invention, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in the present invention and their authorization should be obtained in accordance with relevant laws and regulations through appropriate means.

[0023] As an optional application scenario of this invention, such as Figure 1 As shown, this hydropower optimization scheduling system may include at least one terminal device and at least one server. Figure 1 The system is illustrated in the example, which includes a computer 101, a mobile terminal 102, and a server 103, and the terminal devices such as the computer 101 and the mobile terminal 102 are connected to the server 103 through a network 110.

[0024] Specifically, the terminal device can be a smartphone, tablet, laptop, PDA, desktop computer, game console, smart TV, smart wearable device, in-vehicle terminal, VR (Virtual Reality) device, AR (Augmented Reality) device, etc. Server 103 can be a standalone physical server, a server cluster, a distributed system, or a cloud server providing cloud services. Network 110 can be a wired or wireless network, examples of which include, but are not limited to, the Internet, corporate intranet, local area network, wide area network, mobile communication network, and combinations thereof.

[0025] This invention provides a method for optimizing hydropower scheduling. By integrating sediment-ecological coupling indicators and optimization algorithms for collaborative calculation, it overcomes the technical difficulties of traditional scheduling, such as the separation of water and sediment processes and ecological evaluation, and excessively long calculation time. This method achieves synergistic optimization of cascade hydropower generation benefits, sediment balance, and ecological protection.

[0026] According to an embodiment of the present invention, a method for optimizing hydropower scheduling is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0027] This embodiment provides a method for optimized hydropower scheduling. Figure 2 This is a flowchart of a hydropower optimization scheduling method according to an embodiment of the present invention, such as... Figure 2 As shown, the process includes the following steps: Step S201: Obtain the operational status dataset representing the current operational status of the cascade reservoirs, as well as the inflow forecast data and sediment concentration forecast data for multiple future times, and calculate the water and sediment ecological evaluation index dataset representing the water and sediment ecological status based on the operational status dataset.

[0028] The operational status dataset refers to the collection of data reflecting the actual operational status of a cascade reservoir group at the current moment. It can include water level data, reservoir capacity data, inflow data, outflow data, generator output data, sediment concentration data, and sediment transport rate data for each reservoir. The inflow forecast data refers to the set of results predicting the expected runoff at each inflow section of the cascade reservoir group over multiple future scheduling periods. The sediment concentration forecast data refers to the set of results predicting the sediment concentration in the expected water flow at each inflow section of the cascade reservoir group over multiple future scheduling periods. The water and sediment ecological evaluation index dataset refers to the set of indicators used to comprehensively characterize the status of each river section or ecological control section of the cascade reservoir group in terms of water and sediment transport, riverbed scouring and deposition, and ecological response. It includes the sediment transport intensity index, scouring and deposition risk index, ecological suitability index, and ecological sediment coupling index.

[0029] For example, the sediment transport intensity index can be calculated according to the following formula:

[0030] in, For the first The first section of the river The sediment transport intensity index for each scheduling period is used to characterize the strength of the actual sediment transport rate relative to the baseline sediment transport rate. The actual sediment transport rate, For the first The baseline sediment transport rate for each river section, The STI is a normalization function. The smaller the STI, the weaker the sediment transport capacity and the higher the possibility of sedimentation; the larger the STI, the stronger the sediment transport capacity, which may be accompanied by high sediment content, bottom sediment disturbance and decreased water transparency.

[0031] The actual sediment transport rate can be calculated using the following formula:

[0032] in, Number of sediment particle size groups To control the cross-sectional area of ​​the water passage, For the first Sediment of different particle sizes at cross-sectional location The sand concentration at that location, This represents the cross-sectional flow velocity at the corresponding location.

[0033] The normalization function is defined as follows:

[0034] in, and Determined based on historical measured data, designed sediment transport capacity, or data from typical stable river sections.

[0035] For example, the siltation risk index can be calculated using the following formula:

[0036] in, For the first The first section of the river The siltation risk index for each time period This represents the net siltation volume during that period. This refers to the allowable scouring and sedimentation volume for this river section. For the actual bed surface shear stress, The critical initiation shear stress for riverbed sediment. and The weights for the sedimentation rate and the risk of clean water erosion are respectively determined to satisfy the following conditions. .

[0037] The net flushing volume can be calculated using the following formula:

[0038] in, This represents the net siltation volume during that period. and These represent the input and output sediment transport rates of the river section, respectively. This represents the length of the scheduling period.

[0039] Among these methods, the scouring or sedimentation attribute is preserved by using a scouring / deposition direction variable:

[0040] in, For the direction of siltation and flushing, This represents the net siltation volume during that period. Indicates siltation, Indicates flushing, This indicates a basic balance between scouring and silting.

[0041] For example, the ecological suitability index can be calculated according to the following formula:

[0042]

[0043] in, For the first The ecological control section at the first The comprehensive ecological suitability at different time periods is used to comprehensively evaluate the suitability of flow velocity, water depth, water temperature, dissolved oxygen, sediment concentration, and riverbed sediment for target organisms. The number of ecological suitability evaluation factors. For the first The weight of each ecological factor, For the first Single-factor suitability of each ecological factor.

[0044] The fitness of each single factor can be calculated using the following formula:

[0045] in, The cross-sectional average velocity is... The average water depth, Water temperature Dissolved oxygen concentration, The concentration of sand, The median particle size of the riverbed sediment. This refers to a fitness function determined based on measured data of the target fish or benthic organisms, fitness curves, or expert calibration results. The range of values ​​is A higher value indicates more suitable habitat conditions. The ecological flow guarantee is reflected through minimum ecological flow constraints, expressed as... , Let i be the actual outflow from the i-th reservoir or ecological control section during the t-th scheduling period. The minimum ecological flow threshold for the i-th reservoir or ecological control section during the t-th scheduling period is defined, and the actual flow, velocity, and water depth are used as the basic calculation data for HSI.

[0046] For example, the ecological sediment coupling index can be calculated according to the following formula:

[0047]

[0048]

[0049] in, For the first The river section or ecological control section at the first Eco-sediment coupling index for each time period For ecological suitability index, The sediment transport intensity index. The target sediment transport intensity is adapted to riverbed stability and the target ecological state. As the siltation risk index, The weight of the sediment transport deviation term. The ESI is the weight for the scouring and sedimentation risk item. When the HSI is high, the STI is close to the target sediment transport state, and the SCI is low, the ESI value is high; when the ecological hydraulic conditions appear suitable, but the sediment transport or riverbed scouring and sedimentation state is unsuitable, the ESI value decreases accordingly. This index is used to identify scheduling scenarios where the flow and water depth are suitable, but the sediment load and riverbed conditions are unsuitable.

[0050] Step S202: Combining the operational status dataset, the water and sediment ecological evaluation index dataset, and the water inflow forecast data and sediment concentration forecast data for multiple future times, determine the scheduling strategy population. The scheduling strategy population includes multiple individuals, and each individual corresponds to a scheduling strategy.

[0051] The scheduling strategy population refers to a set of multiple individuals, each corresponding to a candidate scheduling scheme. The population is continuously updated and evolved during the iterative optimization process to approximate the optimal scheduling strategy set. The population size is preset during algorithm initialization. An individual is a basic unit in the population, corresponding to a complete scheduling strategy, and includes the power generation flow data and discharge flow data of each cascade reservoir in each future scheduling period.

[0052] Cascade reservoir operation is a sequential decision-making problem with multiple constraints, multiple objectives, and strong temporal coupling. The operational status dataset provides initial boundary conditions such as current reservoir water level, capacity, and output. The water and sediment ecological evaluation index dataset characterizes the baseline level of current river sediment transport and ecological status. Meanwhile, inflow forecast and sediment concentration forecast data define the external input conditions for future periods. By using these multiple types of data, when generating operation strategies, based on an understanding of the current water and sediment ecological status and future inflow and sediment trends, an initial operation strategy that balances power generation, sediment, and ecological objectives can be generated. This strategy can then be used as the population center to generate the initial population, thereby ensuring that individuals in the initial population are physically feasible and objectively optimal.

[0053] Step S203: Calculate the simulated datasets of water and sediment ecological evaluation indicators and the simulated datasets of operating status at future times after scheduling the current operating status using different individual corresponding scheduling strategies.

[0054] The simulated dataset of water and sediment ecological evaluation indicators refers to the collection of water and sediment ecological evaluation indicators at future time points output after simulating candidate scheduling strategies. This includes simulated values ​​for sediment transport intensity index, scour and deposition risk index, ecological suitability index, and ecological sediment coupling index, used to reflect the changes in water and sediment transport and ecological response after the implementation of the scheduling strategy. The simulated dataset of operational status refers to the collection of operational status data at future time points output after simulating candidate scheduling strategies. This includes simulated water level, outflow, and unit output, used to reflect the changes in the reservoir's operational status after the implementation of the scheduling strategy.

[0055] This process aims to conduct a comprehensive and quantitative evaluation of the effectiveness of each candidate scheduling strategy during iterative optimization. The current operational status alone is insufficient to determine the merits of a scheduling strategy; it is necessary to simulate its future cascading effects, including water level changes, sediment transport, riverbed erosion and deposition, and ecological responses. Only then can we identify non-inferior solutions that achieve a reasonable balance between power generation, sediment balance, and ecological protection. This provides a scientific basis for multi-objective optimization and Pareto ranking, ensuring that the final set of scheduling strategies is not only feasible in the present but also possesses long-term sustainability across the entire water, sediment, and ecological process.

[0056] Step S204: Based on the simulation dataset of water and sediment ecological evaluation indicators and the simulation dataset of operation status, calculate the power generation revenue value, sediment risk value and ecological loss value corresponding to each individual.

[0057] The power generation revenue value refers to the net power generation revenue generated by the dispatch strategy within the dispatch period. It is obtained by subtracting the unit start-up and shutdown costs and water wastage losses from the sum of power generation revenue in each period, reflecting the power generation economic benefits of the dispatch strategy. The sediment risk value is a comprehensive risk measure of the dispatch strategy's impact on river sediment transport and riverbed scour and deposition stability within the dispatch period. It is determined by the degree to which the sediment transport intensity index deviates from the target value in each period and the scour and deposition risk index, reflecting the safety of the dispatch strategy in sediment regulation. The ecological loss value is a comprehensive loss measure of the downstream ecosystem service functions caused by the dispatch strategy within the dispatch period. It is determined by the degree of deficiency of the ecological suitability index and the degree of deficiency of the ecological sediment coupling index in each period, reflecting the comprehensive effect of the dispatch strategy in ecological protection.

[0058] Calculating the power generation revenue, sediment risk, and ecological loss values ​​for each individual is to quantify the performance of each candidate scheduling strategy in the three dimensions of power generation benefit, sediment regulation, and ecological protection into comparable values. This provides basic data for non-dominated ranking in multi-objective optimization. Only by calculating the scores of each individual in the three objectives can we determine whether a scheme is superior to another scheme in all three objectives, and determine which schemes should be retained and which should be discarded. Finally, we output a Pareto optimal scheduling strategy set that takes into account power generation, sediment, and ecology.

[0059] Step S205: Determine the constraint violation amount based on the various types of operating state simulation data and the corresponding constraint boundaries in the operating state simulation dataset.

[0060] Constraint boundaries refer to the permissible range or limits of various operational status data during the operation of cascade reservoirs. These include upper and lower limits of water level, upper and lower limits of outflow, upper and lower limits of generating unit output, ecological flow threshold, and upper limit of scour and sedimentation risk index. Different constraint items can correspond to different constraint boundary values. The ecological flow threshold is dynamically adjusted according to the ecological sensitivity period indicator to determine whether the scheduling strategy meets the basic requirements of safe operation and ecological protection. Constraint violation refers to the total value obtained by normalizing and summing the normalized and squared weighted sum of the excesses of various simulated operational status data of a candidate scheduling strategy relative to its respective constraint boundaries after simulation operation. It is used to quantify the severity of the constraint violation by the scheduling strategy. The larger the constraint violation, the more serious the deviation of the scheduling strategy from the acceptable range in terms of safety or ecological protection.

[0061] For example, the constraint violation amount can be calculated according to the following formula:

[0062] in, To constrain the amount of violation, The weighting coefficient for water level constraints. This is the water level deviation. The weighting coefficients for outbound flow constraints. This is the deviation in outbound flow rate. These are the weighting coefficients for the output constraints. This is the output deviation. The weighting coefficients for ecological flow constraints. This refers to the ecological flow deviation. The weighting coefficients for scouring and silting safety constraints. This represents the deviation amount for siltation risk.

[0063] The water level deviation can be calculated using the following formula:

[0064] in, This is the water level deviation. Let i be the simulated water level of the i-th reservoir or control section during the t-th scheduling period. This represents the highest water level that is allowed to operate at the i-th reservoir or control section. This is the lowest water level that is allowed to operate at the i-th reservoir or control section.

[0065] The ecological flow deviation can be calculated using the following formula:

[0066] in, This refers to the ecological flow deviation. Let be the ecological flow constraint boundary that the i-th reservoir or control section should satisfy during the t-th scheduling period. Let be the simulated outflow from the i-th reservoir or control section during the t-th scheduling period.

[0067] The deviation of scouring and silting risk can be calculated using the following formula:

[0068] in, This is the deviation amount for the risk of siltation. Let be the simulated value of the scouring and sedimentation risk index for the i-th river segment or control section during the t-th scheduling period. The threshold value for the maximum allowable scour and sedimentation risk index for the i-th river segment or control section.

[0069] As an example, a constraint can be represented as: , , , , ,in, This represents the highest water level that is allowed to operate at the i-th reservoir or control section. Let i be the minimum water level that allows operation at the i-th reservoir or control section. Let i be the minimum allowable outflow constraint boundary for the i-th reservoir or control section. Let i be the maximum allowable outflow constraint boundary for the i-th reservoir or control section. Let be the minimum allowable output constraint boundary for the i-th hydropower station unit. This represents the maximum allowable output constraint boundary for the i-th hydropower station unit. Let be the ecological flow constraint boundary that the i-th reservoir or control section should satisfy during the t-th scheduling period. Let i be the simulated outflow from the i-th reservoir or control section during the t-th scheduling period. Let be the simulated value of the scouring and sedimentation risk index for the i-th river segment or control section during the t-th scheduling period. The threshold value for the maximum allowable scour and sedimentation risk index for the i-th river segment or control section.

[0070] Step S206: Determine the optimal scheduling strategy set from the current population based on the constraint violation amount and the power generation revenue value, sediment risk value, and ecological loss value corresponding to each individual.

[0071] The optimal scheduling strategy set refers to the set of all individuals in the current population that are in the first non-dominated layer after non-dominated sorting after the iterative optimization. Each individual in the set corresponds to a scheduling strategy, and they are mutually independent and have their own advantages in the three objectives of power generation revenue, sediment risk, and ecological loss. The set is used for scheduling decision-makers to select a scheduling strategy as the final execution plan according to actual scheduling needs.

[0072] By multiplying the constraint violation by the corresponding penalty coefficient and then summing them to the three objective values, infeasible solutions are penalized on all three objectives and thus ranked in a worse position. Subsequently, the three objective values ​​after penalty correction are non-dominated, so that the individuals in the first non-dominated layer naturally form a Pareto optimal solution set without preset weights. This ensures that the final output scheduling strategy set has theoretical optimality and engineering feasibility on the three objectives of power generation, sedimentation and ecology.

[0073] Step S207: If the optimal scheduling strategy set meets the preset conditions, the cascade reservoirs are scheduled based on the optimal scheduling strategy set.

[0074] The optimal scheduling strategy set satisfies the preset conditions if the current population's optimal scheduling strategy set reaches the preset convergence condition, or if the current iteration count reaches the preset maximum iteration count. For example, the convergence condition is that the change in the optimal scheduling strategy set is less than a preset threshold over several consecutive iterations, meaning the population has converged to a stable Pareto front, or the maximum iteration count is the preset maximum number of iterations, used to forcibly terminate iterations when the algorithm cannot reach the convergence condition.

[0075] This embodiment provides a hydropower optimization scheduling method that constructs a dataset of water and sediment ecological evaluation indicators representing the ecological state of water and sediment, and incorporates it into subsequent calculations. This unifies the sediment transport process and the ecological response process into the scope of scheduling decisions, overcoming the technical defects of traditional scheduling where water and sediment processes and ecological evaluations are separated and it is difficult to capture their dynamic coupling relationship. This achieves synergistic optimization of water and sediment processes and ecological responses. Furthermore, by combining sediment transport deviation, scour and sedimentation risk, ecological suitability, and ecological sediment coordination with constraint violation quantities, the method calculates power generation revenue, sedimentation risk, and ecological loss values, and uses these values ​​for optimization. This ensures that the final output scheduling strategy set can take into account power generation benefits, sediment balance, and ecological protection, providing an intelligent scheduling scheme that takes into account the synergy of multiple water, sediment, and ecological objectives for cascade reservoir scheduling. In addition, by introducing a scour and sedimentation ecological response function to replace the complete physical coupling model for rapid evaluation of candidate schemes, this method overcomes the technical obstacle of excessive computation time and inability to be applied online due to calling the complete water and sediment ecological coupling model, enabling the algorithm to complete large-scale iterative optimization within a limited time.

[0076] In an optional embodiment, step S202, combining the operational status dataset, the water and sediment ecological evaluation index dataset, and the inflow forecast data and sediment concentration forecast data for multiple future times, determines the scheduling strategy population, specifically including: Step a1: Input the operating status dataset, the water and sediment ecological evaluation index dataset, and the water inflow forecast data and sediment concentration forecast data for multiple future times into the preset scheduling strategy model for calculation to obtain the initial scheduling strategy.

[0077] The scheduling strategy model refers to a neural network model that has been trained in advance using a reinforcement learning algorithm, and is used to generate an initial scheduling strategy based on the input state data.

[0078] For example, during the training of the scheduling strategy model, historical operational status datasets, water and sediment ecological evaluation index datasets, historical inflow forecast data, and sediment concentration forecast data are acquired. These data are input into the scheduling strategy model to be trained, yielding the model's output scheduling strategy. This scheduling strategy is then substituted into a pre-set water and sediment ecological coupling model for simulation, obtaining the corresponding simulation results for water and sediment ecological evaluation indices and operational status. Reward values ​​are calculated based on the simulation results, and the parameters of the scheduling strategy model to be trained are updated with the goal of maximizing the cumulative reward until the model converges, completing the training. As an example, the pre-set water and sediment ecological coupling model can also be a pre-trained scouring and deposition ecological response function to reduce the computational load of simulations.

[0079] Step a2: Generate an initial scheduling strategy population using the initial scheduling strategy as the population center individual, and perform optimization calculations on the population to obtain the optimized scheduling strategy population.

[0080] Generating the initial scheduling strategy population by using the initial scheduling strategy as the central individual means taking the initial scheduling strategy as a baseline individual (central individual) in the population. Multiple new individuals are generated around the central individual by adding random perturbations, scaling scheduling strategies based on historical similar operating conditions, etc. The central individual and the multiple new individuals together constitute the initial scheduling strategy population. Optimization computation refers to the iterative process in each generation of the population where the scheduling strategies corresponding to each individual are simulated and evaluated, the objective function value and constraint violation are calculated, and non-dominated sorting is performed. The top-ranked individuals are retained, and a new generation of the population is generated using the whale optimization algorithm based on these retained individuals. This process aims to evolve the population towards better power generation benefits, lower sediment risk, and less ecological loss. The optimized scheduling strategy population refers to the new generation of the population obtained after performing a complete optimization computation on the current population, including simulation evaluation, non-dominated sorting, and individual selection and updating. The optimized population is used to continue participating in optimization computation in the next iteration or as the basis for the final optimal scheduling strategy set output after convergence.

[0081] For example, when generating the population, similar operational processes can be retrieved from the historical scheduling sample database based on the current initial water level, inflow forecast, sediment forecast, and ecological flow requirements. The outflow sequence in the similar scheduling processes is then scaled and perturbed. The reservoir capacity and water level are then calculated time-by-time using the water balance equation. For candidate solutions exceeding the limits of water level, discharge, power output, or ecological flow, constraint projection is used for repair, thereby generating an initial whale population that satisfies physical constraints. This increases the proportion of feasible solutions in the initial population and reduces invalid searches in obviously infeasible regions.

[0082] For example, when population diversity is below a threshold, it can be calculated for some non-elite individuals according to the following formula:

[0083] in, For the new individual generated after applying Gaussian mutation to the j-th individual, For the j-th individual in the current population, there is a corresponding complete scheduling strategy. This is the Gaussian variable asynchrony length coefficient at the k-th iteration, which decreases as the number of iterations increases, and is used to control the amplitude of the variation disturbance. Let be the upper limit vector of the range of values ​​for each decision variable of an individual. Let be the lower bound vector of the value intervals of each decision variable for an individual. For a random vector that follows a standard normal distribution, This is used to generate Gaussian perturbations. Gaussian mutations generate larger perturbations in the early stages to expand the search range, and smaller perturbations in the later stages to improve local search accuracy, thereby preventing individual whales from concentrating too early on the same scheduling scheme.

[0084] For example, a predetermined number of different individuals can be selected from the population to generate differential candidate solutions according to the following formula, where the predetermined number of different individuals can be three:

[0085] And after cross operations, we get:

[0086] in, The difference candidate solution is the mutation vector. Let r1 be the r-th individual randomly selected from the current population. This is the differential evolution scaling factor at the k-th iteration, used to control the scaling factor of the difference vector. Let r2 be the individual randomly selected from the current population. Let r3 be the individual randomly selected from the current population. Let be the candidate value generated for the j-th individual after crossover operations on the d-th dimension decision variable. Let be the variance vector value of the j-th individual on the d-th dimension of the decision variable. Let be the current value of the j-th individual in the current population on the d-th dimension of the decision variable. In the crossover operation, let be the random number generated for the d-th dimension decision variable, with a value range of [0,1]. The crossover probability at the k-th iteration determines the likelihood of the mutated vector value being selected as a candidate individual. The difference vector can utilize the differences between reservoirs and time periods across different scheduling schemes to create new water release processes, thus overcoming the problem that traditional whale optimization algorithms primarily rely on the current optimal individual for their search direction.

[0087] For example, when the improvement in the Pareto front or the optimal target value is less than a preset threshold in several consecutive iterations, the algorithm is considered to have stalled. At this point, the elite individuals on the current Pareto front are retained, and some non-elite individuals are selected for reinitialization. During reinitialization, new individuals are preferentially placed within the action neighborhood recommended by the reinforcement learning strategy, as well as feasible solution regions that the current population has not yet fully explored.

[0088] For example, the selection probability of differential evolution and global search can be increased in the early stages of iteration; and the selection probability of encirclement search, spiral search, and small-scale Gaussian mutation can be increased in the later stages of iteration. The probability of each search operator is dynamically adjusted according to the iteration progress, population diversity, and stagnation state. This can be expressed as:

[0089]

[0090] in, Let be the probability that the differential evolution operator is selected in the k-th iteration. This probability decreases linearly from a maximum to a minimum as the number of iterations increases. It is used to focus on global exploration in the early stages of iteration and local development in the later stages. This represents the maximum probability value of the differential evolution operator being selected, corresponding to the search probability in the initial stage of the iteration. This represents the minimum probability value at which the differential evolution operator is selected, corresponding to the search probability at the end of the iteration. This represents the probability that the Gaussian mutation operator is selected in the k-th iteration. It increases linearly from a minimum to a maximum with the number of iterations, and is used to enhance local search capabilities in the later stages of iteration. This represents the minimum probability that the Gaussian mutation operator will be selected, corresponding to the search probability in the initial stage of the iteration. This represents the maximum probability that the Gaussian mutation operator will be selected, corresponding to the search probability at the end of the iteration. The current iteration number indicates the number of iterations the algorithm has already executed. The maximum number of iterations is preset, and it is one of the termination conditions for the algorithm's iterative optimization.

[0091] This embodiment provides a hydropower optimization scheduling method. By inputting operational status data, water and sediment ecological evaluation index data, and future water and sediment inflow forecast data into a preset scheduling strategy model, the model generates an initial scheduling strategy based on a full understanding of the current water and sediment ecological state and future water and sediment inflow trends. Then, an initial population is generated with this strategy as the central individual and optimization calculations are performed. This overcomes the problems of low feasible solution ratio and poor search efficiency caused by the random generation of the initial population in traditional scheduling optimization. It makes the individuals in the initial population physically feasible and objectively superior, thereby improving optimization efficiency and the quality of the final scheduling strategy set.

[0092] In an optional embodiment, step S203 involves calculating the simulated datasets of water and sediment ecological evaluation indicators and the simulated datasets of operational states for future times after scheduling the current operational state using different individual-specific scheduling strategies. Specifically, this includes: Step b1 involves inputting the current operational status dataset and the scheduling strategies corresponding to different individuals into a preset scouring and sedimentation ecological response function for calculation, thereby obtaining a simulated dataset of water and sediment ecological evaluation indicators and an operational status simulation dataset for future time periods.

[0093] The pre-defined scour-deposition ecological response function refers to a surrogate model trained using historical data, used to replace the complete water-sediment-ecological coupling model. Based on the input current operational status data and candidate scheduling strategies, it quickly predicts simulated data of water-sediment ecological evaluation indicators and operational status at future time points after the implementation of the scheduling strategy. As an example, the scour-deposition ecological response function can be a regression-type response function or a data-driven response function. The simulated dataset of water-sediment ecological evaluation indicators at future time points refers to the set of water-sediment ecological evaluation indicator data output after predicting the candidate scheduling strategy using the scour-deposition ecological response function. This includes simulated values ​​of sediment transport intensity index, scour-deposition risk index, ecological suitability index, and ecological-sediment coupling index at each time point, reflecting the changes in water-sediment transport and ecological response after the implementation of the scheduling strategy. The simulated dataset of operational status at future time points refers to the set of operational status data output after predicting the candidate scheduling strategy using the scour-deposition ecological response function.

[0094] This embodiment provides a hydropower optimization scheduling method that, by inputting the current operating status and different candidate scheduling strategies into a preset scouring and sedimentation ecological response function, can quickly obtain water and sediment ecological evaluation indicators and operating status simulation data at future times after the implementation of each strategy. This overcomes the technical bottleneck of traditional methods, which require too much time to call the complete water and sediment ecological coupling model each time and cannot support large-scale iterative optimization. It provides an efficient computational foundation for the collaborative optimization of reinforcement learning of strategy models and strategy optimization algorithms.

[0095] In an optional embodiment, during step b1, where the current running status dataset and the scheduling strategies corresponding to different individuals are input into a preset scouring and silting ecological response function for calculation, the method for constructing the preset scouring and silting ecological response function specifically includes: Step c1: Obtain multiple training sample data, which include historical running status data and historical scheduling strategy data.

[0096] Training sample data refers to the historical data set used to train the scouring and sedimentation ecological response function, including historical operational status data and historical scheduling strategy data. The historical operational status data consists of operational information such as water level, inflow, outflow, and sediment concentration recorded during the historical operation of the cascade reservoir group; the historical scheduling strategy data consists of the actual power generation flow and discharge flow data of each reservoir within the corresponding historical period.

[0097] For example, after acquiring multiple training sample data, scenario segmentation can be performed on these data. Based on historical scheduling processes, design operation rules, typical inflow years, and ecological scheduling needs, multiple scheduling scenarios can be generated. These scenarios include at least three types of inflow scenarios: high-water years, normal-water years, and low-water years; typical time periods such as flood season, non-flood season, fish spawning migration period, and ecologically sensitive period; and scenarios with different objective preferences such as power generation priority, ecological priority, sediment discharge priority, and comprehensive balance. As an example, each scheduling scenario can be represented as: S = {Qin(t), Z(t), Qout(t), P(t), R(t)}, where Qin(t) is the inflow process, Z(t) is the reservoir water level process, Qout(t) is the outflow process, P(t) is the power plant output process, and R(t) is the scheduling rule or constraint.

[0098] Step c2 involves inputting the training sample data into the water-sediment-ecological coupling model for simulation, obtaining the simulation results of the operational state corresponding to each training sample, and calculating the simulation results of the water-sediment-ecological evaluation index corresponding to each training sample based on the simulation results of the operational state. The water-sediment-ecological coupling model includes a hydrodynamic calculation sub-model, a sediment transport calculation sub-model, a riverbed scour and deposition evolution sub-model, and an ecological response evaluation sub-model. The hydrodynamic calculation sub-model is used to simulate water level, flow velocity, water depth, and flow rate according to the scheduling strategy and boundary conditions. The sediment transport calculation sub-model is used to simulate sediment concentration and sediment transport rate according to the hydrodynamic simulation results. The riverbed scour and deposition evolution sub-model is used to simulate scour and deposition volume and riverbed elevation changes according to the sediment transport simulation results. The ecological response evaluation sub-model is used to calculate the ecological suitability index based on the hydrodynamic simulation results, sediment transport simulation results, and riverbed scour and deposition evolution simulation results.

[0099] The simulation results of the water and sediment ecological evaluation indexes corresponding to each training sample are calculated based on the simulation results of the water level, flow velocity, water depth, sediment concentration, sediment transport rate, scour and deposition volume, and bed shear stress output by the water and sediment ecological coupling model. These simulation results are then substituted into the calculation formulas of sediment transport intensity index, scour and deposition risk index, ecological suitability index, and ecological sediment coupling index to calculate the values ​​of the four core indicators corresponding to each training sample. These values ​​serve as the output labels during the training of the scour and deposition ecological response function.

[0100] The water-sediment-ecological coupling model is a numerical model used to simulate the dynamic changes of hydrodynamics, sediment transport, riverbed scouring and deposition, and ecological response of a cascade reservoir group under a given scheduling strategy. It includes four levels: hydrodynamic calculation sub-model, sediment transport calculation sub-model, riverbed scouring and deposition evolution sub-model, and ecological response evaluation sub-model. The sub-models exchange information through key state variables and are updated synchronously in each scheduling calculation period.

[0101] The hydrodynamic calculation sub-model refers to a numerical model used to simulate the flow process of rivers and reservoirs. It takes the cascade reservoir scheduling scheme as input, and combines the upstream inflow process, inter-section inflow and boundary conditions to calculate the hydrodynamic elements such as flow rate, water level, flow velocity, water depth and water surface slope of each river section. Its output serves as the basic input for the sediment transport calculation sub-model and the ecological response evaluation sub-model.

[0102] The sediment transport calculation sub-model refers to a numerical model used to simulate the sediment transport process in rivers and reservoirs. Based on the hydrodynamic calculation results, it introduces data such as sediment concentration, particle size distribution, sediment inflow process, sediment settling velocity, riverbed composition and reservoir sediment interception and discharge parameters to calculate the sediment flux, sediment concentration changes and sediment transport efficiency of each river section, and establish the sediment transfer relationship between upstream and downstream reservoirs. Its output is used as the input of the riverbed scour and deposition evolution sub-model.

[0103] The riverbed scour and deposition evolution sub-model refers to a numerical model used to simulate changes in riverbed morphology under sediment transport. Based on the differences in sediment inflow, outflow, and transport capacity obtained from the sediment transport calculation sub-model, and combined with data such as river cross-sectional morphology, riverbed elevation, riverbed particle size distribution, roughness, and historical scour and deposition data, it calculates the scour volume, deposition volume, changes in riverbed elevation, and adjustments in channel morphology for each river segment. Its output serves two purposes: firstly, it provides feedback to correct the cross-sectional morphology and roughness parameters in the hydrodynamic calculation sub-model; secondly, it serves as input to the ecological response evaluation sub-model.

[0104] The ecological response evaluation sub-model is an evaluation model used to quantify the impact of scheduling strategies on river ecosystems. It integrates the hydraulic conditions such as flow rate, velocity, and water depth output by the hydrodynamic calculation sub-model, the sediment concentration output by the sediment transport calculation sub-model, and the changes in riverbed elevation and sediment output by the riverbed scouring and deposition evolution sub-model. It calculates the ecological flow satisfaction rate, fish migration suitability, benthic habitat suitability, and comprehensive ecological suitability index, and establishes a quantitative mapping relationship between scheduling methods and ecological responses.

[0105] For example, the output of the hydrodynamic calculation sub-model can be expressed as: H(t) = {Q(t), Z(t), V(t), D(t), J(t)}, where Q(t) is the cross-sectional flow rate, Z(t) is the water level, V(t) is the flow velocity, D(t) is the water depth, and J(t) is the water surface slope. The output of the sediment transport calculation sub-model can be expressed as: M(t) = {C(t), G(t), TSC(t), SE(t), ΔS(t)}, where C(t) is the sediment concentration, G(t) is the sediment flux, TSC(t) is the sediment transport capacity, SE(t) is the sediment transport efficiency, and ΔS(t) is the difference between inflow and outflow sediment. The output of the riverbed scour and deposition evolution sub-model can be expressed as: B(t) = {Ero(t), Dep(t), Δzb(t), BSI(t)}, where Ero(t) is the scour volume, Dep(t) is the deposition volume, Δzb(t) is the riverbed elevation change, and BSI(t) is the riverbed stability index. The output of the ecological response evaluation sub-model can be expressed as: E(t) = {EFR(t), FHSI(t), BHSI(t), TII(t), ARI(t), HSI(t)}, where EFR(t) is the ecological flow satisfaction rate, FHSI(t) is the fish habitat suitability index, BHSI(t) is the benthic habitat suitability index, TII(t) is the transparency impact index, ARI(t) is the algal risk index, and HSI(t) is the comprehensive ecological suitability index.

[0106] Step c3: Using training sample data as input, and the simulation dataset of the running state corresponding to each training sample and the simulation results of water and sediment ecological evaluation indicators as output, the scouring and sedimentation ecological response function to be trained is trained to obtain the trained scouring and sedimentation ecological response function.

[0107] For example, the siltation and sedimentation ecological response function can be a regression response function or a data-driven response function.

[0108] As an example, for indicators with clear physical meaning and relatively stable variable relationships, power functions, exponential functions, or multiple regression functions can be used for fitting, which can be expressed as: Where a, b, c, d, e, and f are parameters to be identified, Q is the cross-sectional flow rate, V is the average cross-sectional velocity, D is the average water depth, C is the sediment concentration, and BSI is the riverbed stability index.

[0109] As an example, for metrics with strong nonlinearity and complex variable coupling, a data-driven response function constructed using random forests, gradient boosting trees, or neural networks can be represented as: Where X represents the scheduling and water and sediment status characteristics, and Y represents response indicators such as scouring and sedimentation volume, ecological suitability, and ecological risk.

[0110] This embodiment provides a hydropower optimization scheduling method that acquires training sample data containing historical operating states and scheduling strategies, and inputs it into a complete water-sediment-ecological coupling model containing four sub-models: hydrodynamics, sediment transport, riverbed scouring and deposition, and ecological response. This generates a high-quality input and output sample library covering various operating conditions. The scouring and deposition ecological response function is then trained based on this sample library. This allows the trained response function to quickly replace the complete physical model, which takes several hours, during the online scheduling phase. This preserves the physical mechanism of water-sediment-ecological coupling while improving computational efficiency.

[0111] In an optional embodiment, the operating status simulation dataset also includes unit output data, unit start-up and shutdown cost data, and water wastage loss data. The water and sediment ecological evaluation index simulation dataset includes a simulated sediment transport intensity index, a simulated scouring and deposition risk index, a simulated ecological suitability index, and a simulated ecological sediment coupling index. Step S204 involves calculating the power generation revenue value, sediment risk value, and ecological loss value for each individual based on the water and sediment ecological evaluation index simulation dataset and the operating status simulation dataset, specifically including: Step d1: Obtain electricity cost data and target sediment transport intensity index for each future time period.

[0112] Electricity cost data refers to the unit price at which the power grid purchases electricity from the hydropower station during each dispatch period, used to calculate power generation revenue for each period. Generator output data refers to the actual electrical power generated by the hydro-turbine generator units during the dispatch period. Generator start-up and shutdown cost data refers to the additional energy consumption losses, mechanical wear and tear, and operation and maintenance costs incurred during the start-up or shutdown of the hydro-turbine generator units, used for deduction when calculating net power generation revenue. Water wastage loss data refers to the loss of power generation benefits caused by the amount of water that is forced to be discharged through spillways or flood discharge facilities when the inflow to the reservoir exceeds the unit's flow capacity and the reservoir's storage capacity, and is used for deduction when calculating net power generation revenue.

[0113] Step d2: Determine the power generation value based on the unit output data and electricity cost data at each time point, and determine the power generation revenue value at each time point by combining the unit start-up and shutdown cost data and water wastage loss data.

[0114] For example, the power generation revenue can be calculated using the following formula:

[0115] in, The total power generation revenue during the dispatch period, The total number of scheduling periods. Let t represent the total number of reservoirs in the cascade reservoir group, t be the scheduling period number, and i be the reservoir number. This is data on electricity costs for a given time period. To contribute to the unit, For unit start-up and shutdown costs, For the loss of water, This represents the length of the scheduling period. The value of electricity generation.

[0116] The power output of the computer group can also be calculated using the following formula:

[0117] in, To contribute to the unit, For power generation and water purification head, For unit efficiency, Let be the power generation flow of the i-th reservoir during the t-th scheduling period.

[0118] Step d3: Determine the simulated sediment transport deviation based on the simulated sediment transport intensity index and the target sediment transport intensity index at each time point, and determine the sediment risk value based on the simulated sediment transport deviation and the simulated scouring and deposition risk index.

[0119] For example, the sediment risk value can be calculated according to the following formula:

[0120] in, The sediment risk value, Let be the sediment transport intensity index for the i-th river segment or control section during the t-th scheduling period, used to characterize the strength of the actual sediment transport rate relative to the baseline sediment transport rate during that period. Let be the target sediment transport intensity index for the i-th river segment or control section during the t-th scheduling period, and be a preset ideal value used to measure the degree of deviation of the actual sediment transport intensity from the ideal state. Let be the scouring and silting risk index for the i-th river segment or control section during the t-th scheduling period. This is the weighting coefficient for the sediment transport deviation in the sediment risk term. This is the weighting coefficient of the erosion and sedimentation risk index in the sediment risk item. To simulate the deviation in sediment transport.

[0121] Step d4: Determine the ecological loss value based on the simulated ecological suitability index and the simulated ecological sediment coupling index.

[0122] For example, the ecological loss value can be calculated according to the following formula:

[0123] in, This represents the ecological loss value. This is the weighting coefficient for the deviation of the ecological suitability index in the ecological loss item. This is the weighting coefficient for the deviation of the ecological sediment coupling index in the ecological loss item. Let be the ecological suitability index of the j-th ecological control section during the t-th scheduling period. Let be the ecological sediment coupling index of the j-th ecological control section during the t-th scheduling period.

[0124] In an optional embodiment, step S206, determining the optimal scheduling strategy set from the current population based on the constraint violation amount and the power generation revenue value, sediment risk value, and ecological loss value corresponding to each individual, specifically includes: Step e1: Determine the penalty amount for power generation revenue, sediment risk, and ecological loss based on the product of the penalty coefficients corresponding to the power generation revenue value, sediment risk value, and ecological loss value and the constraint violation amount.

[0125] The power generation revenue penalty is used to correct the power generation revenue of individuals who violate the constraints, thus penalizing them in terms of power generation targets. The sediment risk penalty is used to correct the sediment risk value of individuals who violate the constraints, thus penalizing them in terms of sediment targets. The ecological loss penalty is used to correct the ecological loss value of individuals who violate the constraints, thus penalizing them in terms of ecological targets.

[0126] Step e2: Determine the power generation revenue value, sediment risk value, and ecological loss value after penalty correction based on the power generation revenue value, sediment risk value, and ecological loss value, as well as the corresponding penalty amount.

[0127] For example, the corrected power generation revenue can be calculated using the following formula:

[0128] in, To penalize the revised power generation revenue value, The net profit after deducting costs. This is the penalty coefficient for the power generation revenue item, and it is a preset value. To constrain violations.

[0129] in, , To constrain the power generation revenue value before the correction for violations.

[0130] For example, the corrected sediment risk value can be calculated according to the following formula:

[0131] in, To penalize the revised sediment risk value, The original sediment risk value, The penalty coefficient for the sediment risk item. To constrain violations.

[0132] For example, the corrected ecological loss value can be calculated according to the following formula:

[0133] in, To penalize the revised ecological loss value, This represents the original ecological loss value. This is the penalty coefficient for the ecological loss item. To constrain violations.

[0134] Step e3: Perform non-dominated sorting on the power generation revenue value, sediment risk value and ecological loss value corresponding to each individual in the current population after penalty correction, and take the first preset number of individuals after sorting as the optimal scheduling strategy set for the current population.

[0135] Non-dominated ranking refers to a method of hierarchically ranking all individuals in a population based on dominance relationships in multi-objective optimization.

[0136] Because the three objectives of power generation revenue, sediment risk, and ecological loss are conflicting and cannot be simultaneously optimized—for example, increasing power generation often leads to increased downstream scouring risk or insufficient ecological flow—a non-dominated ranking method is needed to weigh these three objectives. This method can objectively identify all non-inferior solutions that have advantages in each of the three objectives without pre-setting their weights, thus providing decision-makers with a complete Pareto optimal strategy set and supporting flexible selection as needed.

[0137] For example, a non-dominated ranking of the penalized power generation revenue, sediment risk, and ecological loss values ​​for each individual in the current population can be expressed as follows: .

[0138] This embodiment provides a hydropower optimization scheduling method that multiplies the constraint violation amount by the penalty coefficient corresponding to each objective and then adds it to the power generation revenue value, sediment risk value, and ecological loss value. This ensures that individuals violating the constraints are penalized in all three objectives, thus naturally placing them at a disadvantage in the non-dominated ranking and prioritizing their elimination. There is no need to set a separate feasibility priority rule. At the same time, the non-dominated ranking performs multi-objective trade-offs on the three objective values ​​after penalty correction, so that the selected optimal scheduling strategy set not only meets the basic constraints of engineering safety and ecological protection, but also achieves an objective balance among the three objectives of power generation, sediment, and ecology without the need for preset weights. This ensures the engineering feasibility and multi-objective optimality of the final output strategy.

[0139] This embodiment also provides a hydropower optimization scheduling device, which is used to implement the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0140] This embodiment provides a hydropower optimization scheduling device, such as... Figure 3 As shown, it includes: The data acquisition module 301 is used to acquire the operating status dataset representing the current operating status of the cascade reservoirs, as well as the inflow forecast data and sediment concentration forecast data for multiple future times, and to calculate the water and sediment ecological evaluation index dataset representing the water and sediment ecological status based on the operating status dataset. The strategy calculation module 302 is used to combine the operation status dataset, the water and sediment ecological evaluation index dataset, and the water inflow forecast data and sediment concentration forecast data for multiple future times to determine the scheduling strategy population. The scheduling strategy population includes multiple individuals, and each individual corresponds to a scheduling strategy. The simulation calculation module 303 is used to calculate the simulation dataset of water and sediment ecological evaluation indicators and the simulation dataset of operating status at future times after scheduling the current operating status using different individual corresponding scheduling strategies. The target calculation module 304 is used to calculate the power generation revenue value, sediment risk value and ecological loss value for each individual based on the water and sediment ecological evaluation index simulation dataset and the operation status simulation dataset. The constraint violation calculation module 305 is used to determine the constraint violation amount based on various types of running state simulation data and corresponding constraint boundaries in the running state simulation dataset. The strategy ranking module 306 is used to determine the optimal scheduling strategy set from the current population based on the constraint violation amount and the power generation revenue value, sediment risk value and ecological loss value corresponding to each individual; The scheduling execution module 307 is used to schedule the cascade reservoirs based on the optimal scheduling strategy set if the optimal scheduling strategy set meets the preset conditions.

[0141] The hydropower optimization scheduling device provided in this embodiment of the invention can execute a hydropower optimization scheduling method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects for executing the method. Further functional descriptions of the various modules and units are the same as those in the corresponding embodiments described above, and will not be repeated here.

[0142] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0143] The following is a detailed reference. Figure 4 This diagram illustrates a structural schematic suitable for implementing an electronic device according to embodiments of the present invention. The electronic device may include a processor (e.g., a central processing unit, graphics processor, etc.) 401, which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 402 or a program loaded from memory 408 into random access memory (RAM) 403. The RAM 403 also stores various programs and data required for the operation of the electronic device. The processor 401, ROM 402, and RAM 403 are interconnected via a bus 404. An input / output (I / O) interface 405 is also connected to the bus 404.

[0144] Typically, the following devices can be connected to I / O interface 405: input devices 406 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 407 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; memory devices 408 including, for example, magnetic tapes, hard disks, etc.; and communication devices 409. Communication device 409 allows electronic devices to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 4 Electronic devices with various devices are shown, but it should be understood that it is not required to implement or have all of the devices shown, and more or fewer devices may be implemented or have instead.

[0145] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 409, or installed from a memory 408, or installed from a ROM 402. When the computer program is executed by the processor 401, it performs the functions defined in the hydropower optimization scheduling method of an embodiment of the present invention.

[0146] Figure 4 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of the present invention.

[0147] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, it implements the hydropower optimization scheduling method shown in the above embodiments.

[0148] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0149] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for optimized scheduling of hydropower, characterized in that, The method includes: Obtain the operational status dataset representing the current operational status of the cascade reservoirs, as well as the inflow forecast data and sediment concentration forecast data for multiple future times, and calculate the water and sediment ecological evaluation index dataset representing the water and sediment ecological status based on the operational status dataset. By combining the aforementioned operational status dataset, water and sediment ecological evaluation index dataset, and water inflow forecast data and sediment concentration forecast data for multiple future time periods, a scheduling strategy population is determined. The scheduling strategy population includes multiple individuals, each corresponding to a scheduling strategy. The simulation datasets of water and sediment ecological evaluation indicators and the simulation datasets of operating status are calculated separately for each future time after scheduling the current operating status using different individual corresponding scheduling strategies. Based on the simulated dataset of water and sediment ecological evaluation indicators and the simulated dataset of operational status, the power generation revenue value, sediment risk value and ecological loss value corresponding to each individual are calculated respectively. The constraint violation amount is determined based on the various types of operational state simulation data and the corresponding constraint boundaries in the operational state simulation dataset. The optimal scheduling strategy set is determined from the current population based on the constraint violation amount and the power generation revenue value, sediment risk value, and ecological loss value corresponding to each individual. If the optimal scheduling strategy set meets the preset conditions, the cascade reservoirs are scheduled based on the optimal scheduling strategy set.

2. The method according to claim 1, characterized in that, By combining the aforementioned operational status dataset, water and sediment ecological evaluation index dataset, and forecasts of inflow and sediment concentration at multiple future time points, the scheduling strategy population is determined, including: The operational status dataset, the water and sediment ecological evaluation index dataset, and the water inflow forecast data and sediment concentration forecast data for multiple future times are input into the preset scheduling strategy model for calculation to obtain the initial scheduling strategy. The initial scheduling strategy is used as the population center individual to generate an initial scheduling strategy population, and the population is optimized to obtain an optimized scheduling strategy population.

3. The method according to claim 1, characterized in that, The simulation datasets of water and sediment ecological evaluation indicators and operational status simulation datasets for future time periods are calculated separately after scheduling the current operational status using different individual scheduling strategies, including: The current operating status dataset and the scheduling strategies corresponding to different individuals are input into a preset scouring and sedimentation ecological response function for calculation, so as to obtain a simulated dataset of water and sediment ecological evaluation indicators and a simulated dataset of operating status for each future time.

4. The method according to claim 3, characterized in that, The method for constructing the preset scouring and silting ecological response function includes: Acquire multiple training sample data, including historical running status data and historical scheduling strategy data; The training sample data is input into the water-sediment-ecological coupling model for simulation to obtain the simulation results of the operational state corresponding to each training sample. Based on the simulation results of the operational state, the simulation results of the water-sediment-ecological evaluation index corresponding to each training sample are calculated. The water-sediment-ecological coupling model includes a hydrodynamic calculation sub-model, a sediment transport calculation sub-model, a riverbed scour and deposition evolution sub-model, and an ecological response evaluation sub-model. The hydrodynamic calculation sub-model is used to simulate water level, flow velocity, water depth, and flow rate according to the scheduling strategy and boundary conditions. The sediment transport calculation sub-model is used to simulate sediment concentration and sediment transport rate according to the hydrodynamic simulation results. The riverbed scour and deposition evolution sub-model is used to simulate scour and deposition volume and riverbed elevation changes according to the sediment transport simulation results. The ecological response evaluation sub-model is used to calculate the ecological suitability index based on the hydrodynamic simulation results, sediment transport simulation results, and riverbed scour and deposition evolution simulation results. Using the training sample data as input, and the simulation results of the running state dataset and water and sediment ecological evaluation index corresponding to each training sample as output, the scouring and sedimentation ecological response function to be trained is trained to obtain the trained scouring and sedimentation ecological response function.

5. The method according to claim 1, characterized in that, The operational status simulation dataset also includes unit output data, unit start-up and shutdown cost data, and water wastage loss data. The water and sediment ecological evaluation index simulation dataset includes simulated sediment transport intensity index, simulated scouring and deposition risk index, simulated ecological suitability index, and simulated ecological sediment coupling index. Based on the water and sediment ecological evaluation index simulation dataset and the operational status simulation dataset, the power generation revenue value, sediment risk value, and ecological loss value corresponding to each individual are calculated, including: Obtain electricity cost data and target sediment transport intensity index at various future time points; The power generation output value is determined based on the unit output data and the electricity cost data at each time point, and the power generation revenue value at each time point is determined by combining the unit start-up and shutdown cost data and the water wastage loss data. The simulated sediment transport deviation is determined based on the simulated sediment transport intensity index at each time point and the target sediment transport intensity index, and the sediment risk value is determined based on the simulated sediment transport deviation and the simulated scouring and deposition risk index. The ecological loss value is determined based on the simulated ecological suitability index and the simulated ecological sediment coupling index.

6. The method according to claim 1, characterized in that, Based on the constraint violation amount and the corresponding power generation revenue, sediment risk, and ecological loss value for each individual, the optimal scheduling strategy set is determined from the current population, including: The penalty amount for power generation revenue, the penalty amount for sediment risk, and the penalty amount for ecological loss are determined by multiplying the penalty coefficients corresponding to the power generation revenue value, the sediment risk value, and the ecological loss value with the constraint violation amount. The power generation revenue value, sediment risk value, and ecological loss value, as well as the corresponding penalty amounts, are determined based on the power generation revenue value, sediment risk value, and ecological loss value, after penalty correction. The power generation revenue, sediment risk, and ecological loss values ​​of each individual in the current population are non-dominated and sorted. The first preset number of individuals after sorting are taken as the optimal scheduling strategy set for the current population.

7. A hydropower optimization scheduling device, characterized in that, The device includes: The data acquisition module is used to acquire the operational status dataset representing the current operational status of the cascade reservoirs, as well as the inflow forecast data and sediment concentration forecast data for multiple future times, and to calculate the water and sediment ecological evaluation index dataset representing the water and sediment ecological status based on the operational status dataset. The strategy calculation module is used to combine the operation status dataset, the water and sediment ecological evaluation index dataset, and the water inflow forecast data and sediment concentration forecast data for multiple future times to determine the scheduling strategy population. The scheduling strategy population includes multiple individuals, and each individual corresponds to a scheduling strategy. The simulation calculation module is used to calculate the simulated datasets of water and sediment ecological evaluation indicators and the simulated datasets of operating status at future times after scheduling the current operating status using different individual corresponding scheduling strategies. The target calculation module is used to calculate the power generation revenue value, sediment risk value and ecological loss value for each individual based on the simulated dataset of water and sediment ecological evaluation indicators and the simulated dataset of operating status. The constraint violation calculation module is used to determine the constraint violation amount based on various types of operating state simulation data and corresponding constraint boundaries in the operating state simulation dataset. The strategy ranking module is used to determine the optimal scheduling strategy set from the current population based on the constraint violation amount and the power generation revenue value, sediment risk value and ecological loss value corresponding to each individual. The scheduling execution module is used to schedule the cascade reservoirs based on the optimal scheduling strategy set if the optimal scheduling strategy set meets the preset conditions.

8. An electronic device, characterized in that, include: The system includes a memory and a processor, which are interconnected. The memory stores computer instructions, and the processor executes the computer instructions to perform a hydropower optimization scheduling method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to execute a hydropower optimization scheduling method according to any one of claims 1 to 6.

10. A computer program product, characterized in that, It includes computer instructions for causing a computer to execute a hydropower optimization scheduling method according to any one of claims 1 to 6.

Images