Electric power market transaction-oriented drainage basin cascade water-light complementation optimization scheduling method

Abstract

The invention discloses a drainage basin cascade water-light complementation optimization scheduling method for electric power market transactions, and relates to the technical field related to electric power scheduling, and the method comprises the steps: carrying out the independent power generation fitting modeling of hydraulic power generation and photovoltaic power generation based on environment factors, equipment states and historical data; combining the price fluctuation and demand data of the electricity market to construct a plurality of typical electricity consumption scenes, and fitting to generate different bidding strategies; solving is optimized by taking market income maximization, supply and demand balance, cascade complementary stability and equipment health as targets; and formulating a joint bidding strategy and a corresponding optimal scheduling scheme. The technical problems that in the prior art, scheduling does not fully consider the power market environment, dual uncertainty of photovoltaic output and electricity price is difficult to deal with, and economic benefits are limited due to insufficient water and light resource collaboration are solved, and a complementary strategy of multi-market collaboration and risk optimization is achieved. And the technical effect of improving the overall income and anti-risk capability of watershed cascade water-light complementation in the electricity market is achieved.

Description

Technical Field

[0001] This invention relates to the field of power dispatching technology, and in particular to a basin-wide cascade hydro-solar complementary optimal dispatching method for power market transactions. Background Technology

[0002] As two important forms of renewable energy, cascade hydropower and photovoltaic power generation in a river basin have stable output and strong regulation capabilities, but are affected by hydrological conditions and cascade coupling constraints. Photovoltaic power generation is clean and low-carbon and has widely distributed resources, but its output is intermittent and fluctuates. Complementing the two can mitigate the uncertainty of photovoltaic output to a certain extent and improve the overall system's dispatchability and market competitiveness. In the context of the electricity market, power generation companies need to formulate reasonable bidding strategies based on market price signals and demand fluctuations to maximize profits. However, traditional cascade scheduling in river basins often aims to maximize power generation or meet predetermined loads, without fully considering the price fluctuations and diverse trading instruments in the electricity market environment, such as spot and frequency regulation ancillary service markets. This leads to a disconnect between scheduling decisions and market behavior, resulting in failure to maximize economic benefits and a failure to deeply synergize the rapid regulation capabilities of cascade hydropower with the cost advantages of photovoltaics in terms of time, space, and market dimensions. Existing research on hydro-solar complementary scheduling focuses on physical-level coordination and control, such as compensating for photovoltaic fluctuations through hydropower regulation capabilities, or formulating power generation plans based on deterministic forecasts in day-ahead, intraday / real-time, daily energy, and frequency regulation markets, which is difficult to adapt to changes in the market environment. In addition, cascade hydro-solar complementary scheduling in river basins also faces constraints and challenges such as the spatiotemporal coupling relationship of hydropower between cascade hydropower stations, the randomness of photovoltaic output, the long-term impact of equipment operating health status, and the supply and demand balance requirements in the electricity market.

[0003] At present, the relevant technologies have technical problems such as insufficient consideration of the power market environment in dispatching, difficulty in dealing with the dual uncertainties of photovoltaic output and electricity price, and insufficient synergy between water and photovoltaic resources, which limit economic benefits. Summary of the Invention

[0004] This application provides a basin-wide cascade hydro-solar complementary optimal scheduling method for electricity market transactions. It solves the technical problems in the prior art, such as insufficient consideration of the electricity market environment in scheduling, difficulty in coping with the dual uncertainties of photovoltaic output and electricity price, and limited economic benefits due to insufficient synergy between hydro and solar resources. It achieves the technical effect of improving the overall profitability and risk resistance of basin-wide cascade hydro-solar complementary power in the electricity market through a complementary strategy of multi-market synergy and risk optimization.

[0005] This application provides a basin-wide cascade hydro-solar complementary optimal scheduling method for electricity market transactions. The method includes: after collecting dynamic environmental factors, combining the equipment status and historical power output data of hydropower and photovoltaic power generation equipment, performing independent generation fitting modeling, and establishing independent generation fitting modeling results; acquiring electricity market price fluctuations and electricity demand data, and constructing N typical electricity consumption scenarios based on the price fluctuations and electricity demand data; under the N typical electricity consumption scenarios, performing bidding strategy fitting based on equipment status and real-time price fluctuations, and establishing multiple bidding strategy fitting results; performing basin-wide cascade hydro-solar complementary scheduling analysis based on the independent generation fitting modeling results and the multiple bidding strategy fitting results, wherein the basin-wide cascade hydro-solar complementary scheduling analysis takes market revenue maximization, supply and demand balance constraints, cascade energy complementarity stability constraints, and equipment health constraints as joint objectives, and performs multi-objective optimization solution; configuring a joint bidding strategy based on the multi-objective optimization solution results, and generating an optimized scheduling strategy mapped to the joint bidding strategy.

[0006] In a possible implementation, the basin-wide cascade hydro-solar complementary optimal scheduling method for electricity market transactions further performs the following processing: based on the equipment status and dynamic environmental factors, using the independent generation fitting modeling results to analyze the optimal power generation of hydropower and photovoltaic power generation equipment, and establishing the optimal power generation analysis results; using the optimal power generation analysis results to perform scenario adaptation analysis for N typical electricity consumption scenarios, and establishing an adaptation evaluation degree; using the real-time price fluctuations to establish a price reward and penalty factor; using the independent generation fitting modeling results to perform deviation penalty analysis of the optimal power generation analysis results, and establishing a deviation penalty factor; and using the adaptation evaluation degree, price reward and penalty factor, and deviation penalty factor to fit bidding strategies, and output multiple bidding strategy fitting results.

[0007] In a possible implementation, the basin-wide cascade hydro-solar complementary optimal scheduling method for electricity market transactions further performs the following processing: activating the balance optimization channel, initializing the balance optimization channel using the adaptation evaluation degree, and inputting the price reward / penalty factor and deviation penalty factor into the balance optimization channel; performing bidding strategy fitting by balancing market revenue, equipment health, equipment stability, supply and demand balance, and cascade energy complementarity objectives, and outputting multiple bidding strategy fitting results.

[0008] In a possible implementation, the basin-wide cascade hydro-solar complementary optimal scheduling method for electricity market transactions further performs the following processing: taking each of the fitting results of multiple bidding strategies as an analysis branch and establishing a mapping branch sub-channel; within the mapping branch sub-channel, performing joint objective solution analysis under the corresponding bidding strategy fitting results, and outputting the channel optimization solution results corresponding to each mapping branch sub-channel; and constructing a multi-objective optimization solution result based on the channel optimization solution results.

[0009] In a possible implementation, the basin-level cascade hydro-solar complementary optimal scheduling method for electricity market transactions further performs the following processing: obtaining the joint objective fitness value of each bidding strategy fitting result based on the channel optimization solution results; performing joint weighted analysis using the joint objective fitness value and the strategy priority value of the bidding strategy fitting results to establish a fusion order ranking; and configuring a joint bidding strategy according to the fusion order ranking.

[0010] In a possible implementation, the basin-wide cascade hydro-solar complementary optimal scheduling method for electricity market transactions further performs the following processing: obtaining the optimal ranking result in the fusion order sorting, and using the optimal ranking result as the main bidding strategy; taking the main bidding strategy as the center, calculating the strategy distance between the fitting results of the remaining bidding strategies and the main bidding strategy through a distance metric; adding taboos to the fitting results of bidding strategies whose strategy distance is less than a preset distance threshold, and then selecting the second ranking result in the optimal ranking result that has not been tabulated to construct an auxiliary bidding strategy; and outputting the main bidding strategy and the auxiliary bidding strategy as a joint bidding strategy.

[0011] In a possible implementation, the basin-wide cascade hydro-solar complementary optimal scheduling method for electricity market transactions further performs the following processing: constructing multiple objective functions, including a market revenue maximization objective function, a supply-demand balance objective function, a cascade energy complementarity stability objective function, and an equipment health objective function; weighting and fusing the multiple objective functions to construct a joint objective function; and performing a multi-objective optimization algorithm within the corresponding sub-channel based on the joint objective function to solve the problem and establish the channel optimization solution results.

[0012] In a possible implementation, the basin-wide cascade hydro-solar complementary optimal scheduling method for electricity market transactions further performs the following processing: the cascade energy complementarity stability objective function is as follows: ; in, Characterizing the stability of energy complementarity at different levels, Characterizing time The power generation capacity of hydroelectric power equipment Characterizing time The power generation capacity of photovoltaic power generation equipment Characterizing time The rate of change in the power generation of hydroelectric equipment, Characterizing time The rate of change of power generation of photovoltaic power generation equipment The total power output of the hydro-solar hybrid power generation, Characterizing time The required power, The time interval representing the time step, These are the corresponding weighting coefficients.

[0013] In a possible implementation, the basin-wide cascade hydro-solar complementary optimal scheduling method for electricity market transactions further performs the following processing: obtaining the predicted trust value of dynamic environmental factors; executing the trust identifier of the independent generation fitting modeling result based on the predicted trust value; and performing basin-wide cascade hydro-solar complementary optimal scheduling management based on the independent generation fitting modeling result after the trust identifier.

[0014] In a possible implementation, the basin-wide cascade hydro-solar complementary optimal scheduling method for electricity market transactions further performs the following processing: configuring dynamic update cycle nodes for independent generation fitting modeling results based on the predicted trust value, and re-executing the verification and collection of dynamic environmental factors according to the dynamic update cycle nodes to update the independent generation fitting modeling results.

[0015] This application proposes a basin-wide cascade hydro-solar complementary optimal scheduling method for electricity market transactions. Based on environmental factors, equipment status, and historical data, it independently models the generation of hydropower and photovoltaic power. Combining electricity market price fluctuations and demand data, it constructs multiple typical electricity consumption scenarios and generates different bidding strategies. The method optimizes the solution with the objectives of maximizing market returns, balancing supply and demand, ensuring the stability of cascade complementarity, and maintaining equipment health. Finally, it formulates joint bidding strategies and corresponding optimized scheduling schemes. This method addresses the technical problems in existing technologies, such as insufficient consideration of the electricity market environment, difficulty in handling the dual uncertainties of photovoltaic output and electricity prices, and limited economic benefits due to insufficient synergy between hydropower and photovoltaic resources. It achieves the technical effect of improving the overall returns and risk resistance of basin-wide cascade hydro-solar complementary power in the electricity market through multi-market collaboration and risk optimization complementary strategies. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings of the embodiments of the present invention will be briefly described below. Flowcharts are used in this application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed precisely in sequence. Instead, various steps can be processed in reverse order or simultaneously as needed. Furthermore, other operations can be added to these processes, or one or more steps can be removed from these processes.

[0017] Figure 1 This is a flowchart illustrating a basin-wide cascade hydro-solar complementary optimal scheduling method for electricity market transactions, provided as an embodiment of this application.

[0018] Figure 2 This is a flowchart illustrating the process of establishing the fitting results of multiple bidding strategies in a watershed cascade hydro-solar complementary optimal scheduling method for electricity market transactions, as provided in an embodiment of this application. Detailed Implementation

[0019] To further illustrate the technical means and effects adopted by the present invention in order to achieve the intended purpose, the following detailed description is provided in conjunction with the accompanying drawings and preferred embodiments, based on the specific implementation methods, structures, features and effects of the present invention.

[0020] This application provides a method for optimized scheduling of cascade hydro-solar hybrid power systems in a river basin, oriented towards electricity market transactions. Figure 1 As shown, the method includes: Step S100: After collecting dynamic environmental factors, combine the equipment status and historical power output data of hydropower equipment and photovoltaic power generation equipment to perform independent power generation fitting modeling and establish independent power generation fitting modeling results.

[0021] Preferably, the independent power generation fitting modeling involves establishing models for each hydropower station and photovoltaic power station within the basin, capable of accurately predicting their future power generation based on environmental factors and equipment status. Dynamic environmental factors are collected; for hydropower, this primarily involves hydrological and meteorological data, such as basin precipitation, reservoir inflow, current reservoir water level, and evaporation, which directly determine the available water resources for the hydropower station. For photovoltaic power generation, this mainly involves meteorological data, such as solar irradiance, ambient temperature, cloud cover, and humidity, which directly determine the power generation efficiency of the photovoltaic panels. The equipment status of hydropower and photovoltaic power generation equipment refers to the health and operational limitations of the equipment itself, such as the number of available turbine generator units and their start-stop status, unit efficiency curves, whether there are maintenance plans or output limitations, the health of equipment such as water intake pipelines and gates, the availability of photovoltaic modules, the operating status and conversion efficiency of inverters, and the health status of equipment such as combiner boxes and transformers. Historical power output data refers to the past actual power generation records of hydropower and photovoltaic power stations, as well as historical environmental data corresponding to the actual power generation records, such as historical flow and historical irradiance.

[0022] Preferably, for hydropower, independent power generation fitting modeling employs a physical model based on the turbine's comprehensive efficiency curve and head-flow characteristics. This model, combined with historical data, uses least squares or support vector regression to fit parameters, establishing a functional relationship between head, power generation flow, and output power. Specifically, a theoretical model is obtained by establishing a relationship based on the turbine characteristic curve and generator efficiency formula. This model is then calibrated and fitted using historical power output data. For example, regression models, random forests, and neural networks are used to learn how to more accurately predict the final power generation under complex operating conditions, based on inputs such as inflow, current water level, and unit combination. For photovoltaic power generation, independent power generation fitting modeling employs a single-diode equivalent circuit model or a gradient boosting tree model using irradiance and module temperature as inputs. This model establishes a functional relationship between meteorological conditions and output power. Specifically, a basic model is obtained by establishing a relationship based on the physical characteristics of photovoltaic modules. This model is also fitted using historical power output data. A machine learning model is used to learn the actual output power of the power station under specific irradiance, ambient temperature, and module temperature conditions, resulting in an empirical model that better reflects actual site conditions and includes factors such as line loss and dust obstruction. Finally, the independent power generation fitting modeling results are output, including the power generation prediction model for hydropower stations and the power generation prediction model for photovoltaic power stations.

[0023] Step S200: Obtain price fluctuations and electricity demand data in the electricity market, and construct N typical electricity consumption scenarios based on the price fluctuations and electricity demand data.

[0024] Preferably, quantitative analysis is performed on the uncertain future prices and loads in the electricity market, generating a set of probabilistic future market states to provide a comprehensive risk environment simulation. Specifically, this involves acquiring electricity market price fluctuation data, including historical price data, projected price data, and related factors. Historical price data refers to the time-of-use clearing prices in the electricity spot market over the past few months or years, for example, one price point every 15 minutes or hour. Projected price data is a point prediction and uncertainty estimate of electricity prices for a specific future period based on market supply and demand analysis, fuel costs, and renewable energy output forecasts. Related factors are data on drivers affecting prices, such as historical load, weather information, and holiday types. Electricity demand data is also acquired, including historical demand data, projected demand data, and related factors. Historical demand data is the total system load curve for the same time period in the past; projected demand data is a prediction of the total system load for a specific future period and its uncertainty estimate; and related factors are data on drivers affecting load, such as date type, temperature, and economic activity index.

[0025] Preferably, future electricity prices and loads are continuous and uncertain random variables. By analyzing historical data and using time series models to perform joint probability distribution analysis, joint dynamic characteristics are identified. For example, high temperatures may lead to increased load and higher electricity prices, which is a positive correlation. Then, based on the determined joint probability distribution, multiple possible future 24-hour electricity price and load sequences are randomly generated through Monte Carlo simulations, etc., where each sequence is called an original scenario. The original scenarios are then refined using k-means clustering, forward selection, and backward reduction, supplementing the objective function of scenario reduction, such as minimizing the probability distance between the original scenario set and the typical scenario set. N typical scenarios are used to represent the overall statistical characteristics of the original large set of scenarios, and each typical scenario is assigned a specific objective function. The probability weight represents the likelihood of a scenario occurring. N is a positive integer representing the number of typical scenarios. The final output consists of N structured typical electricity consumption scenarios, including the predicted electricity price and predicted electricity demand / load for each scenario, as well as the corresponding scenario probability. The sum of the probabilities of all typical electricity consumption scenarios is 1. For example, Scenario 1 is a sunny workday with high photovoltaic output, medium load, and medium electricity price, with a probability of 35%. Scenario 2 is a hot workday with high photovoltaic output but the efficiency may decrease slightly in the afternoon due to high temperatures, extremely high load, and extremely high electricity price, with a probability of 20%. Scenario 3 is a cloudy or rainy workday with low photovoltaic output, slightly different load due to insufficient sunlight and commercial activities, and higher electricity price due to lack of sunlight, with a probability of 25%. Scenario 4 is a holiday with low load, low electricity price, and potentially high photovoltaic output but low absorption value, with a probability of 20%.

[0026] Step S300: Under N typical electricity consumption scenarios, the bidding strategy is fitted based on equipment status and real-time price fluctuations, and multiple bidding strategy fitting results are established.

[0027] Furthermore, such as Figure 2 As shown, step S300 further includes step S310, which involves analyzing the optimal power generation of hydropower and photovoltaic power generation equipment based on the equipment status and dynamic environmental factors using the independent power generation fitting modeling results, and establishing the optimal power generation analysis results; step S320, which involves using the optimal power generation analysis results to perform scenario adaptation analysis for N typical electricity consumption scenarios, and establishing an adaptation evaluation degree; step S330, which involves using the real-time price fluctuations to establish a price reward and penalty factor; step S340, which involves using the independent power generation fitting modeling results to perform deviation penalty analysis on the optimal power generation analysis results, and establishing a deviation penalty factor; and step S350, which involves fitting bidding strategies based on the adaptation evaluation degree, price reward and penalty factor, and deviation penalty factor, and outputting multiple bidding strategy fitting results.

[0028] Preferably, the optimal power generation analysis of hydropower and photovoltaic power generation equipment is conducted by using the independent power generation fitting modeling results, equipment status, and dynamic environmental factors as inputs. The optimal power generation analysis determines the theoretical maximum power generation capacity curves of hydropower and photovoltaic power generation within the dispatch cycle, without considering market factors and only considering equipment capacity and the natural environment. For photovoltaics, predicted irradiance is directly input into its fitting model to obtain the maximum possible power generation curve, i.e., the maximum output of photovoltaics under current natural conditions. For hydropower, considering cascade hydraulic connections and reservoir capacity constraints, the maximum total electricity or its power distribution that the hydropower system can generate within the dispatch cycle, while satisfying physical constraints such as water balance and reservoir capacity limitations, is determined. The final output is the optimal power generation analysis results, which include multiple time-series curves representing the upper limit of power generation capacity of water and electricity at the natural-physical level.

[0029] Preferably, the optimal power generation analysis results are used to conduct scenario adaptation analysis for N typical electricity consumption scenarios. This involves quantitatively evaluating the degree to which the temporal characteristics of power generation, based on the optimal power generation analysis results, matches market demand in each typical market electricity consumption scenario. Specifically, using the optimal power generation analysis results and N typical electricity consumption scenarios containing multiple load and price curves as input, multiple matching indicators are calculated for each typical electricity consumption scenario. For example, the correlation or fitting error between the combined power generation and the predicted load curve in the scenario is calculated to determine load tracking capability; the better the power generation curve "follows" the load curve, the higher the adaptation degree. The calculation of whether power generation capacity is high during the periods with the highest electricity prices determines peak coverage. Then, by combining load tracking capability and peak coverage, an adaptation evaluation degree is calculated for each typical electricity consumption scenario. A higher adaptation evaluation degree indicates a better match between the natural power generation characteristics and market demand in that market environment. The adaptation evaluation degree is used to quantitatively evaluate the shape matching degree between the power generation curve and the market load curve and the coverage capability for high-value periods in a specific market scenario. The calculation formula is as follows: , Let represent the fit evaluation score for scenario s, with a value range of [0, 1]. A higher score indicates a more significant advantage in that scenario. , Let be the weighting coefficient, satisfying + =1, The normalized shape correlation coefficient is given by the formula: , This represents the maximum theoretical power generation at time t. For the predicted load at time t in scenario s, , They are respectively and The mean value within the scheduling period T. The peak-hour value coverage rate is calculated using the following formula: , The set of peak electricity price periods defined in scenario s is used to measure the proportion of the system's power generation capacity that meets market demand during high electricity price periods. The closer the ratio is to or greater than 1, the better the system can seize high-yield opportunities.

[0030] Preferably, a price reward / penalty factor is established using real-time price fluctuations, directly quantifying market price signals into incentive or penalty coefficients for power generation behavior. This price reward / penalty factor is positively correlated with electricity prices; that is, generating electricity during periods of high electricity prices yields higher reward scores. Then, deviation penalty analysis is performed on the optimal power generation analysis results using independent power generation fitting modeling results to quantitatively assess the risk and cost of actual power generation deviating from the bidding plan. Specifically, for photovoltaic power generation, its output is highly volatile, and the prediction confidence value is low on cloudy days, resulting in a high deviation risk and thus a higher deviation penalty factor. For hydropower generation, its output is highly controllable, and the prediction confidence value is high, resulting in a low deviation risk and a low deviation penalty factor. The deviation penalty factor represents the risk cost of the actual output not matching the bid amount due to inaccurate predictions. Finally, a bidding strategy fitting process is performed, which integrates the fit evaluation degree, price reward / penalty factor, and deviation penalty factor to initially generate several different feasible bidding strategies. These strategies represent the planned power generation for each time period in the market. For example, an aggressive profit-oriented strategy schedules hydropower output during high-price periods as much as possible, ignoring some stability issues and pursuing high fit evaluation degree and high price reward / penalty factor, but may incur a higher deviation penalty factor. A robust conservative strategy smooths the combined output curve, closely following the optimal power generation analysis results to minimize the deviation penalty factor, but may sacrifice some high-price revenue. A load-tracking strategy prioritizes matching the bidding curve with the predicted load curve to obtain a high fit evaluation degree, which may have an advantage in the ancillary power service market. Using the fit evaluation degree, price reward / penalty factor, and deviation penalty factor as objective constraints, the strategy space is quickly searched to generate a Pareto optimal solution set that balances revenue and risk. This set is then output as the fitting result of multiple bidding strategies to improve the efficiency of the entire dispatch decision-making process.

[0031] Preferably, the price reward / penalty factor is used to directly embed market price signals into the optimization objective. Its value is proportional to the electricity price, aiming to incentivize the system to generate more electricity during periods of high electricity prices and to generate less electricity or provide ancillary services during periods of low electricity prices. ; in, Let be the price reward / penalty factor at time t in scenario s. For the predicted electricity price at time t in scenario s, The benchmark electricity price is used as the basis for calculation. A positive reward is generated when the electricity price exceeds this benchmark, and a negative penalty is generated when it falls below it. For reference electricity prices, used for normalization, for example, the standard deviation of historical electricity prices can be used. , As a sensitivity coefficient for rewards and punishments, and set This makes the system more sensitive to "price increases" than to "price decreases," encouraging a more aggressive pursuit of high profits and reflecting an asymmetric market response strategy.

[0032] Preferably, the deviation penalty factor is used to quantify the risk cost of actual output deviating from the bidding plan, and is positively correlated with the uncertainty of its output forecast. The calculation formula is as follows: ,in, It is a type of power generation equipment. Deviation penalty factor It is a type of power generation equipment. The variance of the power output prediction error is derived from the statistical analysis of historical prediction data and actual power generation data, and can be dynamically adjusted by the prediction confidence value. When the confidence value is low, a larger variance is assigned. Estimate; for photovoltaics, Strongly correlated with weather type and forecast duration, for hydropower, The value is relatively small, but increases during flood season or unit failure; loss of load value It is a relatively large constant representing the socioeconomic cost of power shortages; risk discount factor. This is an adjustment coefficient related to the market deviation assessment rules. For example, if the market's penalty for deviation is linear, then... It can be set as a penalty rate; if it's a secondary penalty, then... Corresponding coefficients.

[0033] Furthermore, step S350 also includes step S351, activating the balance optimization channel, performing balance optimization channel initialization using the adaptation evaluation degree, and inputting the price reward / penalty factor and deviation penalty factor into the balance optimization channel; step S352, performing bidding strategy fitting by balancing market revenue, equipment health, equipment stability, supply and demand balance, and tiered energy complementarity objectives, and outputting multiple bidding strategy fitting results.

[0034] Preferably, an optimization solution unit is set up as a balanced optimization channel to perform multi-objective trade-off decisions to generate the final bidding strategy. Activating the balanced optimization channel means allocating computing resources to the optimization solution unit and loading relevant configuration parameters. Then, the balanced optimization channel is initialized using the fit evaluation degree, including setting the initial search direction and setting initial weight values. For example, the exploration is prioritized in the strategy space corresponding to scenarios with high fit evaluation degree. Initial weights are assigned to the optimization objectives of different scenarios, with scenarios with high fit evaluation degree having an advantage in initial weights. Then, market and economic driving parameters such as price reward / penalty factors and deviation penalty factors are input into the balanced optimization channel as key coefficients and constraints of the optimization objectives, respectively. Then, with the objectives of balancing market returns, equipment health, equipment stability, supply and demand balance, and tiered energy complementarity, the bidding strategy fitting is performed. That is, multi-objective optimization is carried out to determine the Pareto optimal solution set and output multiple bidding strategy fitting results. Each strategy is a different trade-off. For example, strategy A has extremely high market returns, but slightly poor equipment health score and acceptable stability; strategy B has moderate market returns, but the best equipment health score and very stable output; strategy C has balanced indicators, with no particularly outstanding features or obvious weaknesses.

[0035] Preferably, market revenue is driven by price incentive factors, with the goal of maximizing total electricity sales revenue; equipment health typically translates into operational constraints, such as avoiding frequent start-ups and shutdowns of hydro turbines in vibration zones and reducing overload operation of photovoltaic inverters to extend equipment lifespan; equipment stability refers to the smoothness of output from individual devices, avoiding drastic power fluctuations that could trigger equipment protection actions; supply-demand balance refers to the degree of matching between the total power generation plan and the total load forecast, to meet the stringent requirements for safe grid operation; cascade energy complementarity stability is system-level stability, ensuring coordinated output between hydropower and photovoltaics to smooth the total output power and avoid impacting the grid.

[0036] Step S400: Based on the independent power generation fitting modeling results and the fitting results of multiple bidding strategies, a basin-wide cascade hydro-solar complementary scheduling analysis is performed. The basin-wide cascade hydro-solar complementary scheduling analysis takes market revenue maximization, supply and demand balance constraints, cascade energy complementarity stability constraints, and equipment health constraints as joint objectives, and performs multi-objective optimization solution.

[0037] Preferably, the basin-wide cascade hydro-solar complementary scheduling analysis is conducted based on the independent power generation fitting modeling results and the fitting results of multiple bidding strategies. This involves finding the Pareto front among the fitting results of multiple bidding strategies through multi-objective optimization. Specifically, the basin-wide cascade hydro-solar complementary scheduling analysis uses market revenue maximization, supply-demand balance constraints, cascade energy complementarity stability constraints, and equipment health constraints as joint objectives to perform multi-objective optimization. Market revenue maximization directly includes the revenue items from the electricity market and ancillary service market, with the objective of maximizing them. The supply-demand balance constraint is a grid security requirement, requiring the system's total power generation plan to be equal to the load forecast, serving as a hard constraint. The cascade energy complementarity stability constraint is a constraint that the power output must be below a certain threshold to ensure the smoothness and stability of the hydro-solar synergy. Equipment health constraints are long-term operation and asset maintenance requirements, such as constraining turbines to avoid vibration zones, limiting the number of daily start-ups and shutdowns of units, and controlling the temperature of photovoltaic inverters within a safe range. The joint objective refers to assigning weights to multiple objectives and performing a weighted summation to obtain a total objective function. This function is then used for multi-objective optimization. All objectives, constraints, input data, and fitting results of multiple bidding strategies are used as the initial population or initial solution to input into the total objective function. The output is the final, executable scheduling scheme in Pareto optimization, which specifies the physical scheduling instructions and market bidding strategies, namely the power generation of each power station in each time period and the power volume and frequency regulation capacity to be bid in the market for this power generation plan. This ensures a deep closed-loop integration of power market transactions and physical operation.

[0038] Furthermore, step S400 also includes step S410, taking each of the multiple bidding strategy fitting results as an analysis branch and establishing a mapping branch sub-channel; step S420, within the mapping branch sub-channel, performing joint objective solution analysis under the corresponding bidding strategy fitting result, and outputting the channel optimization solution result corresponding to each mapping branch sub-channel; step S430, constructing a multi-objective optimization solution result based on the channel optimization solution result.

[0039] Preferably, each bid strategy fitting result among multiple bid strategy fitting results is treated as an analysis branch. That is, an independent optimization solution unit or computation thread / process is created for each bid strategy fitting result to achieve logical partitioning of computing resources, resulting in multiple mapped branch sub-channels. Each branch sub-channel corresponds one-to-one with the initial bid strategy, which is loaded into the corresponding branch sub-channel as its initial solution or core parameters for optimization. Then, within the mapped branch sub-channel, a joint objective solution analysis under the corresponding bid strategy fitting results is executed. Specifically, the input bid strategy fitting result is used as the core guide for the initial point or search direction, and a complete optimization solution process considering market revenue, supply and demand balance, equipment stability, and equipment health is run. Since each branch sub-channel starts from a different bid strategy, its optimization path and final local optimum are also different, that is, it explores the final goal from multiple directions simultaneously. After completing the optimization calculation, each branch sub-channel outputs the optimal scheduling scheme and the corresponding objective function score as the channel optimization solution result. Finally, the channel optimization results of all parallel branch sub-channels are combined to form a candidate solution set, and a multi-objective optimization solution is constructed to clearly show the best trade-offs that may exist among multiple competing objectives, ensuring that a high-quality Pareto optimal solution set is found more quickly.

[0040] Furthermore, step S420 also includes step S421, constructing multiple objective functions, including a market revenue maximization objective function, a supply and demand balance objective function, a cascade energy complementarity stability objective function, and an equipment health objective function, and weighting and fusing the multiple objective functions to construct a joint objective function; step S422, executing a multi-objective optimization algorithm within the corresponding sub-channel based on the joint objective function to solve the problem and establishing the channel optimization solution result.

[0041] Step S421 further includes the following objective function for the cascade energy complementarity stability: ; in, Characterizing the stability of energy complementarity at different levels, Characterizing time The power generation capacity of hydroelectric power equipment Characterizing time The power generation capacity of photovoltaic power generation equipment Characterizing time The rate of change in the power generation of hydroelectric equipment, Characterizing time The rate of change of power generation of photovoltaic power generation equipment The total power output of the hydro-solar hybrid power generation, Characterizing time The required power, The time interval representing the time step, These are the corresponding weighting coefficients. Through mathematical construction, an optimized model is established to coordinate and complement hydropower and photovoltaic output in a time-series manner. The first term minimizes the absolute difference in instantaneous output between hydropower and photovoltaics, ensuring they form a complementary "one rises, the other falls" pattern at any given time, avoiding simultaneous high or low power generation. The second term minimizes the difference in their power change rates, constraining hydropower output changes to track photovoltaic fluctuations, thus smoothing the ramp-up process of combined output and preventing sudden power fluctuations from impacting the grid. The third term minimizes the deviation between total power generation and system demand, ensuring that the overall output after complementarity can effectively track the load curve and meet the grid's supply and demand balance requirements. After the three terms are weighted and integrated, the system will automatically optimize the compensation strategy for photovoltaic fluctuations by hydropower, while meeting power demand, achieving stable, smooth, and dispatchable combined power output.

[0042] Preferably, multiple objective functions are constructed, including a market revenue maximization objective function, a supply-demand balance objective function, a cascade energy complementarity stability objective function, and an equipment health objective function. The market revenue maximization objective function uses market price and bid power as inputs to calculate total revenue in the electricity market and ancillary service market. The supply-demand balance objective function uses total hydropower and solar power generation and system load or bid power as inputs to calculate the total deviation between power generation and demand, minimizing this value to meet the basic requirements for safe grid operation. The cascade energy complementarity stability objective function includes a power difference term and a power change rate difference term, utilizing the differences to form a stable total output and penalize uncoordinated ramp-up, respectively. The equipment health objective function quantifies equipment wear and fatigue, penalizing frequent start-ups and shutdowns of units. Alternatively, it could penalize turbines by operating them in inefficient or vibration-prone areas. Then, multiple objective functions are weighted and fused to construct a joint objective function. This involves assigning weights to each objective based on decision-making preferences, directly determining scheduling tendencies. For example, a high weight for the market revenue maximization objective indicates revenue priority; even if equipment wears out more or output fluctuates more, revenue will be increased. Finally, a multi-objective optimization algorithm is executed based on the joint objective function. Calculations are performed within each branch sub-channel, searching for the power generation plan that maximizes the joint objective function value while satisfying all physical and market constraints. The final output is the channel optimization solution, i.e., the local optimum within each branch sub-channel, including the optimal power generation plan, the corresponding bidding strategy, and the values ​​of each objective function under that plan.

[0043] Furthermore, step S430 also includes step S431, obtaining the joint objective fitness value of each bidding strategy fitting result based on the channel optimization solution result; step S432, performing joint weighted analysis using the joint objective fitness value and the strategy priority value of the bidding strategy fitting result to establish a fusion order ranking; and step S433, configuring the joint bidding strategy according to the fusion order ranking.

[0044] Preferably, the original values ​​of each objective are normalized based on the channel optimization solution to eliminate the influence of dimensions. Then, corresponding weights are assigned to each objective based on historical decision preference data, for example, a profit weight of 0.35, a stability weight of 0.3, an equipment health weight of 0.2, and a supply-demand balance weight of 0.15. A weighted comprehensive calculation is then used to quantitatively evaluate the performance of each candidate solution, obtaining a joint objective fitness value. The higher the value, the better the solution performs under comprehensive consideration. This yields the joint objective fitness value for each bidding strategy fitting result. Then, a joint weighted analysis is performed using the joint objective fitness value and the strategy priority value of the bidding strategy fitting result. The strategy priority value is a set preference coefficient representing a preference for a certain type of bidding strategy. For example, if the current... Given the higher market risk, the initial priority value of a conservative strategy is set higher. If the current strategy is to maximize profits, the priority value of an aggressive profit-generating strategy will be higher. The strategy priority value is determined based on historical data learning or manually set by the dispatcher to obtain a final score. The weighting coefficient is used to balance objective performance and subjective preferences. Then, all candidate schemes are ranked from high to low according to the final score, i.e., the fusion order ranking. The joint bidding strategy is configured according to the fusion order ranking. That is, the scheme ranked first in the fusion order ranking is directly used as the primary bidding strategy, and schemes with less similarity in the ranking are excluded. The top few schemes are then used as auxiliary or backup strategies. Finally, the joint bidding strategy is output, including the generation plan and bidding volume curve used to report to the electricity market.

[0045] Furthermore, step S433 also includes: obtaining the optimal ranking result in the fusion order sorting, and using the optimal ranking result as the main bidding strategy; taking the main bidding strategy as the center, calculating the strategy distance between the fitting results of the remaining bidding strategies and the main bidding strategy through a distance metric; adding taboos to the fitting results of bidding strategies whose strategy distance is less than a preset distance threshold, and then selecting the second ranking results in the optimal ranking results that have not been tabulated to construct an auxiliary bidding strategy; and outputting the main bidding strategy and the auxiliary bidding strategy as a joint bidding strategy.

[0046] Preferably, the scheme ranked first in the fusion order is directly selected as the optimal ranking result and used as the primary bidding strategy, representing the core plan reported to the electricity market and used to guide production. Then, with the primary bidding strategy as the center, the strategy distance between the fitting results of the remaining bidding strategies and the primary bidding strategy is calculated using distance metrics. That is, with the primary bidding strategy as the benchmark, the strategy distance between each of the remaining strategies in the ranking list and the primary bidding strategy is calculated. The square root of the sum of the squares of the differences between the two power curves at each time point is used to measure the similarity between the two strategies. The smaller the distance, the more similar they are. A preset distance threshold is set as the boundary for strategies that are too similar. All bidding strategies with a strategy distance less than the preset distance threshold are marked as taboo. Then, among the remaining strategies that are not taboo, the second-highest ranked result in the initial fusion order is selected as the auxiliary bidding strategy to provide valuable and alternative alternative paths. Finally, the primary bidding strategy and the auxiliary bidding strategy are output as a joint bidding strategy, thereby ensuring the ability to cope with market changes and sudden failures and improving the overall dispatch resilience.

[0047] Step S500: Configure the joint bidding strategy based on the multi-objective optimization solution results, and generate an optimized scheduling strategy mapped to the joint bidding strategy.

[0048] Preferably, configuring a joint bidding strategy based on the multi-objective optimization solution results refers to decomposing and packaging the power generation plan included in the multi-objective optimization results according to market rules, formatting it into bidding documents required by the electricity market, for example, directly or in segments converting the power plan curve into a time-of-use electricity bidding curve and filling it into the market's standard bidding form; generating a frequency regulation capacity bidding curve based on the capacity that can be used for frequency regulation determined in the optimization results, and finally generating a joint bidding strategy to be reported to the power trading center. The market bidding decision is then decomposed into production instructions for each specific power station and unit, serving as corresponding optimized scheduling strategies. For example, a joint bidding strategy might specify a total output of 500MW from 10:00 to 10:15, with the optimized scheduling strategy mapping this to 150MW output for Unit 1 of Hydropower Station X, 100MW output for Unit 2 of Hydropower Station X, 200MW output for Photovoltaic Power Station Y, and 50MW output for Hydropower Station Z, specifying their reservoir discharge flow. This generates detailed, time-series in-plant scheduling commands, which are then distributed to the monitoring systems of each power station and the reservoir scheduling system. These commands include, but are not limited to, unit start / stop orders, target output setpoints, gate opening instructions, and reservoir water level control targets. This ensures that the optimal economic decisions made in the market are executed accurately by the physical system, truly achieving a deep integration of market transactions and physical operations.

[0049] Furthermore, the basin-wide cascade hydro-solar complementary optimal scheduling method for electricity market transactions also includes: obtaining the predicted trust value of dynamic environmental factors; executing the trust identifier of the independent generation fitting modeling result based on the predicted trust value; and performing basin-wide cascade hydro-solar complementary optimal scheduling management based on the independent generation fitting modeling result after the trust identifier.

[0050] Preferably, the prediction confidence value of dynamic environmental factors is calculated based on historical performance and predictive science principles. The prediction confidence value is a value between 0 (completely unreliable) and 1 (completely reliable). Specifically, it involves continuously comparing historical prediction data with actual observation data. For example, for photovoltaic prediction, under a continuous cloudy weather pattern, the variance of its prediction error increases, thus lowering the prediction confidence value under this weather pattern. For hydrological prediction, when the watershed is experiencing periods of drastic change such as heavy rain or floods, the input uncertainty of the hydrological model increases, and its output prediction confidence value also decreases. Conversely, during periods of normal water levels, the confidence value is higher. For example, the confidence value for a 24-hour photovoltaic prediction under clear and stable high-pressure weather conditions might be 0.95; however, under spring weather conditions with alternating cold and warm air currents and rapidly moving clouds, the confidence value might only be 0.70. Then, based on the predicted confidence value, a confidence flag is applied to the independent generation fitting modeling results. This means the predicted confidence value is input as a key parameter into the optimized scheduling model. The optimized scheduling management of the basin-wide cascade hydro-solar complementary power generation is then performed based on the independent generation fitting modeling results after the confidence flag is applied. If the confidence value is high, the prediction is considered highly reliable, and a more proactive and aggressive strategy is adopted. This involves setting a smaller deviation penalty factor in the optimized model, allocating more hydropower resources for arbitrage, and maximizing power generation during peak electricity price periods. If the confidence value is low, the prediction is considered highly uncertain, and a more conservative and robust strategy is adopted. This involves increasing the deviation penalty factor, tending to allow hydropower stations to reserve more standby capacity, and also reducing aggressive bidding during high-risk periods or increasing reserved capacity in the balanced market. This enables the entire scheduling system to dynamically adapt to uncertain environments, significantly improving its practicality and robustness in the real world.

[0051] Furthermore, the basin-wide cascade hydro-solar complementary optimal scheduling method for electricity market transactions also includes configuring a dynamic update cycle node for the independent generation fitting modeling results based on the predicted trust value, and re-executing the verification and collection of dynamic environmental factors according to the dynamic update cycle node to update the independent generation fitting modeling results.

[0052] Preferably, the dynamic update cycle nodes for the independent power generation fitting modeling results are configured based on the prediction confidence value. Specifically, when the confidence value is low, it is judged that the current performance of the model may be poor or the environment is changing rapidly, so the update cycle is shortened to allow it to catch up with the environmental changes as soon as possible and restore prediction accuracy. When the confidence value is high, it is judged that the model performance is stable and the environmental changes are gradual, so the update cycle is extended to save computing resources and maintain model stability. Then, the verification and collection of dynamic environmental factors are re-executed according to the dynamic update cycle nodes. That is, dynamic environmental factor data are collected in real time within the configured dynamic update cycle nodes, such as the actual solar irradiance and temperature in the past few hours, the actual inflow and rainfall in the past few hours, and the actual power generation data of the corresponding time period. The real-time collected dynamic environmental data is merged with historical data to form an updated dataset, and then the fitting models of hydropower and photovoltaic power generation are retrained and fine-tuned to generate more accurate independent power generation fitting modeling results. This ensures a better description of the current and future power generation capacity and improves the overall benefits and risk resistance of the basin's cascade hydropower-photovoltaic complementarity in the power market.

[0053] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A basin-wide cascade hydro-solar complementary optimal scheduling method for electricity market transactions, characterized in that: The method includes: After collecting dynamic environmental factors, and combining the equipment status and historical power output data of hydropower and photovoltaic power generation equipment, independent power generation fitting modeling is performed to establish the independent power generation fitting modeling results. Acquire electricity market price fluctuations and electricity demand data, and construct N typical electricity consumption scenarios based on the price fluctuations and electricity demand data; In N typical electricity consumption scenarios, bidding strategies are fitted based on equipment status and real-time price fluctuations, and multiple bidding strategy fitting results are established. Based on the independent power generation fitting modeling results and the fitting results of multiple bidding strategies, a basin-level cascade hydro-solar complementary scheduling analysis is performed. The basin-level cascade hydro-solar complementary scheduling analysis takes market revenue maximization, supply and demand balance constraints, cascade energy complementarity stability constraints, and equipment health constraints as joint objectives, and performs multi-objective optimization solution. Configure a joint bidding strategy based on the multi-objective optimization solution results, and generate an optimized scheduling strategy that maps to the joint bidding strategy.

2. The basin-wide cascade hydro-solar complementary optimal scheduling method for electricity market transactions as described in claim 1, characterized in that, Under N typical electricity consumption scenarios, bidding strategies are fitted based on equipment status and real-time price fluctuations, and multiple bidding strategy fitting results are established, including: Based on the equipment status and dynamic environmental factors, the optimal power generation of hydropower and photovoltaic power generation equipment is analyzed using the independent power generation fitting modeling results, and the optimal power generation analysis results are established. Using the optimal power generation analysis results, scenario adaptation analysis is performed on N typical power consumption scenarios to establish an adaptation evaluation degree; A price reward / penalty factor is established using the aforementioned real-time price fluctuations; Using the independent power generation fitting modeling results, deviation penalty analysis is performed on the optimal power generation analysis results to establish a deviation penalty factor; Based on the aforementioned fit evaluation degree, price reward and punishment factor, and deviation penalty factor, the bidding strategy is fitted, and multiple bidding strategy fitting results are output.

3. The basin-wide cascade hydro-solar complementary optimal scheduling method for electricity market transactions as described in claim 2, characterized in that, Based on the aforementioned fit evaluation degree, price reward / penalty factor, and deviation penalty factor, a bidding strategy is fitted, and multiple bidding strategy fitting results are output, including: Activate the balance optimization channel, and after performing balance optimization channel initialization using the adaptation evaluation degree, input the price reward / penalty factor and deviation penalty factor into the balance optimization channel; By balancing market returns, equipment health, equipment stability, supply and demand balance, and tiered energy complementarity objectives, the bidding strategy is fitted, and multiple bidding strategy fitting results are output.

4. The basin-wide cascade hydro-solar complementary optimal scheduling method for electricity market transactions as described in claim 1, characterized in that, Based on the independent power generation fitting modeling results and the fitting results of multiple bidding strategies, a basin-wide cascade hydro-solar complementary scheduling analysis is conducted, including: Each bid strategy fitting result in the multiple bid strategy fitting results is taken as an analysis branch, and a mapping branch sub-channel is established; Within the mapped branch sub-channel, perform joint objective solution analysis under the corresponding bidding strategy fitting results, and output the channel optimization solution results corresponding to each mapped branch sub-channel; Based on the channel optimization solution results, construct the multi-objective optimization solution results.

5. A basin-wide cascade hydro-solar complementary optimal scheduling method for electricity market transactions as described in claim 4, characterized in that, Configure a joint bidding strategy based on the multi-objective optimization solution results, including: The joint objective fitness value of the fitting results for each bidding strategy is obtained based on the channel optimization solution results; A joint weighted analysis is performed using the joint target fitness value and the strategy priority value of the bidding strategy fitting result to establish a fusion order ranking; The joint bidding strategy is configured according to the fusion order.

6. The basin-wide cascade hydro-solar complementary optimal scheduling method for electricity market transactions as described in claim 5, characterized in that, The joint bidding strategy is configured according to the fusion order, including: Obtain the optimal sorting result in the fusion order sorting, and use the optimal sorting result as the main bidding strategy; Centered on the main bidding strategy, the strategy distance between the fitting results of the remaining bidding strategies and the main bidding strategy is calculated using a distance metric. After adding taboos to the fitting results of bidding strategies whose strategy distance is less than a preset distance threshold, the second ranking result that has not been added to the optimal ranking result is selected to construct an auxiliary bidding strategy. The main bidding strategy and the auxiliary bidding strategy are output as a joint bidding strategy.

7. A basin-wide cascade hydro-solar complementary optimal scheduling method for electricity market transactions as described in claim 4, characterized in that, Perform joint objective solution analysis based on the fitting results of the corresponding bidding strategy, including: Multiple objective functions are constructed, including a market return maximization objective function, a supply and demand balance objective function, a cascade energy complementarity stability objective function, and an equipment health objective function. These multiple objective functions are then weighted and fused to construct a joint objective function. Based on the joint objective function, a multi-objective optimization algorithm is executed within the corresponding sub-channel to solve the problem and establish the channel optimization solution results.

8. A basin-wide cascade hydro-solar complementary optimal scheduling method for electricity market transactions as described in claim 7, characterized in that, The objective function for the stability of the cascaded energy complementarity is as follows: ； in, Characterizing the stability of energy complementarity at different levels, Characterizing time The power generation capacity of hydroelectric power equipment Characterizing time The power generation capacity of photovoltaic power generation equipment Characterizing time The rate of change in the power generation of hydroelectric power equipment, Characterizing time The rate of change of power generation of photovoltaic power generation equipment The total power output of the hydro-solar hybrid power generation, Characterizing time The required power, The time interval representing the time step, These are the corresponding weighting coefficients.

9. A basin-wide cascade hydro-solar complementary optimal scheduling method for electricity market transactions as described in claim 1, characterized in that, The results of establishing independent power generation fitting modeling also include: Obtain the prediction confidence value of dynamic environmental factors; Based on the predicted trust value, the trust identifier of the independent power generation fitting modeling result is executed, and the basin-wide cascade hydro-solar complementary optimal scheduling management is carried out based on the independent power generation fitting modeling result after the trust identifier.

10. A basin-wide cascade hydro-solar complementary optimal scheduling method for electricity market transactions as described in claim 9, characterized in that, Based on the predicted trust value, configure a dynamic update cycle node for the independent power generation fitting modeling results, and re-execute the verification and collection of dynamic environmental factors according to the dynamic update cycle node to update the independent power generation fitting modeling results.

Images