Method for optimizing and scheduling multi-region spot market bidding strategy through joint participation of water, wind and light

Abstract

The invention belongs to the technical field of power dispatching, and particularly relates to a multi-region spot market bidding strategy optimization and dispatching method with water-wind-light combined participation, which comprises the following steps: acquiring original power generation data of a hydropower station and a wind-light power generation unit, generating a typical scene of the hydropower station and a typical scene of the wind-light power generation unit through a Latin hypercube sampling method and a clustering algorithm, and generating a typical scene of the hydropower station and a typical scene of the wind-light power generation unit; a typical scene set is obtained through pairwise combination, and a water-wind-light joint participation multi-region spot market optimization bidding double-layer model based on the Stackelberg game is constructed; the water-wind-light system bidding strategy optimization model takes the maximum total income as an objective function, and the multi-region spot market optimization clearing model takes the bid-winning power as a decision variable and takes the minimum electricity purchase cost as an objective; and solving the double-layer model through a distributed iterative algorithm to obtain bid-winning power and clearing electricity price of the water-wind-light system in the spot market of each region, and finally constructing a water-wind-light joint optimization scheduling model in an electricity market environment to obtain a scheduling plan.

Description

Technical Field

[0001] This invention belongs to the field of power market dispatching technology, specifically involving an optimization strategy and dispatching method for joint participation of hydropower, wind power, and solar power in multi-regional spot market bidding. Background Technology

[0002] Against the backdrop of energy transition, clean energy sources, represented by wind and solar power, are developing rapidly. However, the randomness, volatility, and intermittency of their output pose significant challenges to the operation of the electricity spot market, which is centered on real-time balance. Currently, wind and solar power generators face severe deviation assessment risks when participating in the spot market due to limitations in forecast accuracy, resulting in weak market competitiveness. They often have to participate with conservative strategies, making it difficult to truly reflect their value and achieve efficient consumption through market mechanisms. To enhance the market viability of renewable energy, integrating it physically and operationally with cascade hydropower, which is flexible in regulation and has energy storage characteristics, to form an integrated hydro-wind-solar system, and utilizing the rapid regulation capability of hydropower to smooth out wind and solar fluctuations and reduce overall output deviations, has become an important development direction.

[0003] However, when participating in complex transactions across provinces and multiple spot markets, this hydro-wind-solar system faces multiple intertwined challenges in internal optimization and external bidding. First, there is a multi-stakeholder interest game within the system. Although hydropower and wind / solar power are physically complementary, as independent market entities or asset owners, their cost structures, risk preferences, and profit objectives differ, leading to complex trade-offs and cooperative game relationships in joint bidding decisions. Second, there are complex hydraulic-electric coupling constraints within the system. The upstream and downstream power stations in a cascade hydropower system have close hydraulic connections and power output linkages. Their bidding decisions in their respective regional markets must strictly meet the constraints of hydropower time-series balance and reservoir scheduling; otherwise, mismatches in bid output can easily lead to water wastage or forced purchase of electricity at high prices in the real-time market, resulting in economic losses. Third, the system faces the challenge of strategic interactions across multiple markets externally. The spot market prices in different regions vary and are uncertain in time and space. The system needs to formulate a joint bidding strategy in multiple markets and time periods in order to capture arbitrage opportunities and maximize overall returns. This is a complex problem involving high-dimensional and multi-stage decision-making.

[0004] Most existing technical methods fail to simultaneously and systematically address the aforementioned three core issues. They either focus on optimizing a single type of power source, consider only a single market environment, simplify the system to a single decision-making entity while ignoring internal game dynamics, or fail to deeply integrate cross-market strategies with refined cascade hydraulic coupling models. This makes it difficult for existing methods to guide hydro-wind-solar systems in achieving technically feasible, economically optimal, and interest-coordinated bidding strategies in real-world multi-regional spot markets.

[0005] Therefore, there is an urgent need for a comprehensive method that can accurately characterize the multi-agent game behavior within the system, deeply integrate the dynamic constraints of the cascade hydraulic-electric coupling, and effectively optimize the bidding strategy of spot markets across multiple regions. Summary of the Invention

[0006] The purpose of this invention is to provide an optimization and scheduling method for bidding strategies in multi-regional spot markets involving water, wind, and solar power, thereby optimizing the bidding strategies of water, wind, and solar power systems in multi-regional spot markets, coordinating the clearing process of multi-regional spot markets, and improving the spot market revenue of water, wind, and solar power systems.

[0007] To achieve the above objectives, this invention provides an optimization method for multi-regional spot market bidding strategies involving water, wind, and solar power, comprising the following steps: S1. Collect raw power generation data of hydropower station and wind and solar power generation unit at different times. Then, generate typical scenarios of hydropower station and typical scenarios of wind and solar power generation unit by Latin hypercube sampling method and clustering algorithm. By combining them in pairs, obtain S typical scenario sets. The typical scenarios include power generation data change curves over time. S2. Construct a two-layer model for optimizing bidding in the multi-regional spot market based on a multi-leader, multi-follower Stackelberg game involving water, wind, and solar power. The two-layer model includes a bidding strategy optimization model for water, wind, and solar power systems and an optimization clearing model for the multi-regional spot market. The bidding strategy optimization model for water, wind, and solar power systems takes maximizing total revenue as its objective function, while the optimization clearing model for the multi-regional spot market takes the winning bid power as the decision variable and minimizes the electricity purchase cost as its objective. S3. Solve the two-layer model using a distributed iterative solution algorithm to obtain the winning bid power and clearing price of the hydro-wind-solar system in the spot market of various regions.

[0008] Furthermore, in step S1, the generation of the typical scenario specifically includes: S11. Divide the interval [0,1] of the cumulative probability distribution function into... N Divide the sample into equal parts, and then obtain the prediction error for each sample n based on the cumulative probability distribution function. Then, based on the original power generation data and prediction error, the power generation data sampling value is calculated to generate the uncertain scenario of hydropower station and wind and solar power generation system; S12. The uncertain scenarios are clustered using the K-means clustering method to generate typical scenarios for hydropower stations and wind and solar power generation units, and the probability of occurrence of each typical scenario is calculated. S13. Use Cartesian products to combine the typical scenarios of the hydropower station and wind and solar power generation units in pairs to obtain S typical scenario sets, and calculate the occurrence probability of each typical scenario set according to the occurrence probability of the typical scenarios.

[0009] Furthermore, in step S1, the original power generation data of the hydropower station includes the hydropower station runoff variation curve over time, and the original power generation data of the wind and solar power generation system includes the wind and solar power generation power variation curve over time.

[0010] Furthermore, in step S2, the objective function of the bidding strategy optimization model for the water-wind-solar system is as follows:

[0011]

[0012]

[0013] In the formula, Represents water, wind, and light system i In the spot market, I represents the total number of water, wind, and solar systems. Indicating a water-wind-solar system i exist s In the scene t Always d The spot market revenue obtained by a region, where D represents the total number of regions and T represents the total duration of spot market power supply. Represents water, wind, and light system i exist s In the scene t The cost of generating electricity at any given time This represents the probability of scenario s occurring. Represents water, wind, and light system i In the scene s Down t Always in the region d The winning bid power, Represents water, wind, and light system i Zhongshui Hydropower Station g exist s In the scene t The planned power generation at any given time, where G represents the total number of hydropower stations in the water-wind-solar system i. This represents the clearing electricity price in region d at time t under scenario s. The power generation cost coefficient of hydropower station g in the water-wind-solar system i. This represents the power generation of wind and solar power system j within the water-wind-solar system i at time s. The power generation cost coefficient represents the power generation cost of wind and solar power system j within the water-wind-solar system i.

[0014] Furthermore, the optimization model for the bidding strategy of the hydro-wind-solar system includes bidding power constraints, cascade hydropower station operation constraints, and hydropower station output constraints; the bidding power constraints are as follows:

[0015] In the formula, Represents water, wind, and light system i exist s In the scene t Always towards d The bidding power submitted by the region; The operational constraints of the cascade hydropower stations are as follows:

[0016] In the formula, , and Representing water, wind, and solar systems respectively i middle g hydroelectric power station t Water level at the end of the time period and its upper and lower limits; Indicating a water-wind-solar system i middle g The reservoir water level-capacity curve function of a hydroelectric power station; and These represent the initial water level constraint and the final water level constraint, respectively. , and Representing water, wind, and solar systems respectively i middle g hydroelectric power station t The discharge flow rate and its upper and lower limits for each time period; , and Representing water, wind, and solar systems respectively i middle g Hydropower station t The power generation flow rate at any given time and its upper and lower limits; , Representing water, wind, and solar systems respectively i middle g Hydropower station t Storage capacity at time t1 and time t-1; The time interval is 1 hour. R i,g,s,t For water, wind and light system i middle g Hydropower station t Inbound traffic at any given time; I i,g,s,t For water, wind and light system i middle g Hydropower station t Real-time outbound flow; Water, wind and light system i middle g -1 The outflow from the hydropower station reached the first g The time for each hydroelectric power station; The power output constraints of the hydropower station are as follows:

[0017] In the formula, Indicating a water-wind-solar system i middle g Water consumption rate of hydropower stations; and Representing water, wind, and solar systems respectively i middle g Hydropower station t Maximum and minimum power generation within the time period; The bidding curve consists of the bidding power and the generation cost, as shown below:

[0018]

[0019]

[0020] in, Represents water, wind, and light system i In the scene s Down t Always towards d The spot bidding curve submitted by the region, The variable representing the determining coefficient of the linear term of the power bid by the water-wind-solar system i for region d at time t in scenario s is... The variable representing the determining coefficient of the quadratic term of the power bid by the water-wind-solar system i for region d at time t in scenario s is... For the coefficient variable initial value, For the coefficient variable initial value, For the coefficient variable The slope of the change in declared power. For the coefficient variable The slope of the change in declared power.

[0021] Furthermore, the objective function of the multi-regional spot market optimization clearing model is as follows:

[0022] In the formula, The total cost of electricity purchase in the regional spot market. The curve representing the spot bidding price submitted by the water-wind-solar system i to region d at time t under scenario s. for s In the scene t Time Zone d The winning bid power of virtual power plant v, where V is the total number of virtual power plants in region d. Let i be the winning bid power of the hydro-wind-solar power generation system i in scenario s at time t in region d. The spot bidding curve for virtual power plant v located in region d at time t in scenario s; And set regional spot market supply and demand balance constraints and bidding power constraints, as shown below:

[0023]

[0024]

[0025]

[0026] In the formula, for t Time Zone d The spot market demand power, Water, wind and light system i To the region d The capacity of the transmission lines.

[0027] Furthermore, in step S3, the iterative solution of the two-layer model includes the following steps: S31. According to the optimization requirements of the two-level model, input the unit parameters, predicted runoff, predicted power generation, and market demand at each stage for each market bidding participant. S32. Randomly generate initial decision variables for each optimization model in the multi-leader layer, and set the convergence error. And set the iteration count to iter=1, and then start the iteration; S33. Solve the optimization models of the multi-leader layer using the solver to obtain the optimal bidding curve of each market bidding participant in the iter round; S34. Based on the optimal bidding curve obtained in step S33, solve the spot market optimization clearing model for each region in the multi-follower layer, obtain the clearing result of each region in the iter round, and return the result to the multi-leader layer. S35. Starting from the +1th iteration, each market bidding participant in the multi-leader layer updates the bidding curve of the +1th iteration based on the clearing result of the +1st iteration, and inputs the bidding curve into the multi-follower layer to solve for the clearing result of the +1st iteration. S36. Determine whether the solution result of the (iter+1)th round is consistent with the result of the (iter)th round. If they are consistent, the game has reached Nash equilibrium, and the solution result of the (iter+1)th round is the equilibrium solution of the Stackelberg game model. Terminate the iteration. Otherwise, update the iteration count to (iter+2)th and go to step S33 to continue solving until the solution result converges or the maximum number of iterations is reached.

[0028] Furthermore, the iterative solution is optimized using a quadratic update penalty term and a damped update method; the quadratic update penalty term is shown below:

[0029] In the formula, This represents the objective function after adding the penalty term. Represents the original objective function. The decision variable representing the m-th water-wind-solar system bidding strategy optimization model in this round is... represents the decision variable of the previous round's m-th water-wind-solar system bidding strategy optimization model. The smoothing coefficient is used to characterize the adjustment friction resistance of decision variables during the iterative adjustment process, and M represents the total number of bidding strategy optimization models for water, wind and solar systems. The damping update method includes: after obtaining the clearing results of each region in the (iter+1)th round, implementing damping update on the clearing results to ensure that the variables of the upper and lower level models converge towards the new solution. The specific update form is as follows:

[0030] in, This represents the clearing result of the (iter+1)th round after the update. This represents the clearing result of the iter round. This represents the actual clearing result of round (iter+1). This represents the damping coefficient, used to adjust the degree to which a new solution permeates historical strategies.

[0031] The present invention also provides a method for optimizing the scheduling of a multi-regional spot market involving hydropower, wind power, and solar power, comprising: constructing a hydropower, wind power, and solar power joint optimization scheduling model under the power market environment based on the winning bid power obtained by the bidding strategy optimization method for multi-regional spot market involving hydropower, wind power, and solar power as described above, so as to obtain the scheduling plan for each hydropower station in each hydropower, wind power, and solar power system.

[0032] Furthermore, the objective function of the joint water, wind, and solar energy optimization scheduling model is as follows:

[0033] The hydraulic constraints of the cascade hydropower project are shown below:

[0034] The power output constraints of the hydropower station are as follows:

[0035] The contract power balance constraints are as follows:

[0036] in, Let represent the contract power of the water-wind-solar system i at time t in region d.

[0037] In summary, compared with the prior art, the above-described technical solutions conceived by this invention mainly possess the following technical advantages: 1. The multi-regional spot market bidding strategy optimization and scheduling method provided by this invention can accurately characterize the multi-entity game behavior within the system, deeply integrate the dynamic constraints of cascade hydropower-electric coupling, and effectively optimize the bidding strategy across multiple regions, thereby improving the spot market revenue of the hydropower-wind-solar system. It provides scientific decision support for the hydropower-wind-solar system to maximize overall revenue and minimize risk in complex market environments, thus effectively promoting the market-based consumption and optimized allocation of clean energy on a larger scale.

[0038] 2. This invention utilizes a quadratic update penalty term, where any allocation that deviates significantly from the previous round's decision will incur additional penalty costs, thereby forcing the leader to adopt a smoother strategy to adjust the path during iteration. This mechanism effectively enhances the strong convexity of the objective function, making the decision solution more numerically stable, and suppressing directional jumps that may be induced by excessive competition at the game theory level.

[0039] 3. This invention, through a damped update method, effectively suppresses overreactions in strategies caused by competitive behavior, preventing abrupt changes in power and price trajectories during iteration, thereby significantly improving the numerical stability of the game-solving process. Theoretically, damped update can be viewed as a weighted average of the strategy mapping, making it more contractile, thus enhancing the existence and uniqueness of equilibrium points and accelerating the solution speed of the two-layer model. Attached Figure Description

[0040] Figure 1 This is a flowchart illustrating the multi-regional spot market bidding strategy optimization method based on multi-agent game theory involving water, wind, and solar power, provided by the present invention.

[0041] Figure 2 This is a schematic diagram of the solution process for the multi-regional spot market bidding strategy optimization method based on multi-agent game theory with joint participation of water, wind and solar power, provided by the present invention. Detailed Implementation

[0042] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0043] Please see Figure 1-2 This invention provides an optimization method for a multi-regional spot market bidding strategy involving water, wind, and solar power, comprising the following steps: S1. Collect raw power generation data of hydropower station and wind and solar power generation unit at different times. Then, generate typical scenarios of hydropower station and typical scenarios of wind and solar power generation unit by Latin hypercube sampling method and clustering algorithm. By combining them in pairs, obtain S typical scenario sets. The typical scenarios include power generation data change curves over time. S2. Construct a two-layer model for optimizing bidding in multi-regional spot markets based on a multi-leader, multi-follower Stackelberg game involving water, wind, and solar power. The two-layer model includes a bidding strategy optimization model for water, wind, and solar power systems and an optimization clearing model for multi-regional spot markets. The bidding strategy optimization model for water, wind, and solar power systems takes maximizing total revenue as its objective function, while the optimization clearing model for multi-regional spot markets takes the winning bid power as the decision variable and minimizes the electricity purchase cost as its objective. The objective function is calculated based on a set of typical scenarios and the sum of their probabilities. S3. Solve the two-layer model using a distributed iterative solution algorithm to obtain the winning bid power and clearing price of the hydro-wind-solar system in the spot market of various regions.

[0044] The spot market in this invention typically refers to electricity trading over a one-day period.

[0045] Specifically, step S1 includes: The prediction errors of hydropower station runoff and wind and solar power output are generated by Ding Chao cubic sampling method and K-means clustering algorithm, and the uncertain scenario set of runoff and wind and solar power output is obtained based on this.

[0046] 1) Generation of uncertain scenarios for hydropower station runoff and wind and solar power output based on hypercube sampling method set up N The cumulative probability distribution function of the prediction error, representing the number of samples, is: The sampling process is as follows: First, the interval [0,1] of the cumulative probability distribution function is divided into... N Divide the data into equal parts; then, randomly select a point in each sub-interval. θ n,t As prediction error x n,t Sampling points; finally, take Inverse function calculation x n,t As shown in the following formula:

[0047] In the formula, Let be the prediction error of sample n at time t.

[0048] Compared with the original forecast data Multiplying them yields the final sample value. ,as follows:

[0049] Each sample value's curve over time constitutes an uncertain scenario. By performing the above calculations on the original prediction data of hydropower stations and wind and solar power generation units respectively, several uncertain scenarios for hydropower stations and wind and solar power generation units are obtained.

[0050] 2) Generation of typical scenarios for hydropower station runoff and wind and solar power output based on K-means clustering algorithm The K-means clustering method was used to cluster the uncertain scenarios of hydropower stations and wind and solar power generation units respectively, generating two sets of typical scenarios (sample value change curves over time), and the probability of occurrence of each scenario was calculated, which is equal to the ratio of the scenario in that class to the total number of scenarios.

[0051] 3) Scenario using Cartesian product combination of hydropower station runoff and VRE output The Cartesian product is used to combine the typical scenarios of each random variable (power generation data of hydropower stations and wind and solar power units) after K-means classification in pairs to obtain the final typical scenario set (each typical scenario set contains one typical scenario of a hydropower station and one typical scenario of a wind and solar power unit). The total number of combined typical scenario sets is a multiple of the original two scenario sets, and the probability of each scenario is the product of multiples of the probabilities of the original scenarios.

[0052] Step S2 is as follows: 1) Optimization Model for Bidding Strategy of Water-Wind-Solar System Water-wind-solar system i The objective function is as follows:

[0053]

[0054]

[0055] In the formula, Represents water, wind, and light system i In the spot market, I represents the total number of water, wind, and solar systems. Indicating a water-wind-solar system i exist s In the scene t Always d The spot market revenue obtained by a region, where D represents the total number of regions and T represents the total duration of spot market power supply. Represents water, wind, and light system i exist s In the scene tThe cost of generating electricity at any given time This represents the probability of scenario s occurring. Represents water, wind, and light system i In the scene s Down t Always in the region d The winning bid power, Represents water, wind, and light system i Zhongshui Hydropower Station g exist s In the scene t The planned power generation at any given time, where G represents the total number of hydropower stations in the water-wind-solar system i. This represents the clearing electricity price in region d at time t under scenario s. The power generation cost coefficient of hydropower station g in the water-wind-solar system i. This represents the power generation of wind and solar power system j within the water-wind-solar system i at time s. The power generation cost coefficient represents the power generation cost of wind and solar power system j within the water-wind-solar system i.

[0056] Water-wind-solar system i The bidding power constraints are as follows:

[0057]

[0058] in, Represents water, wind, and light system i exist s In the scene t Available bidding power at any given time Represents water, wind, and light system i exist s In the scene t Always towards d The bidding power of regional applications.

[0059] Water-wind-solar system i The output constraints of the cascade hydropower stations are as follows: The hydraulic constraints are as follows:

[0060] In the formula , and Representing water, wind, and solar systems respectively i middle g hydroelectric power station t Water level at the end of the time period and its upper and lower limits; Indicating a water-wind-solar system i middle g The reservoir water level-capacity curve function of a hydroelectric power station; and These represent the initial water level constraint and the final water level constraint, respectively. , and Representing water, wind, and solar systems respectively i middle g hydroelectric power station t The discharge flow rate and its upper and lower limits for each time period; , and Representing water, wind, and solar systems respectively i middle g Hydropower station t The discharge flow rate at any given time and its upper and lower limits; Indicating a water-wind-solar system i middle g Hydropower station t Storage capacity at time t1 and time t-1; The time interval is 1 hour. R i,g,s,t Water, wind and light system i middle g Hydropower station t Inbound traffic at any given time; I i,g,s,t Water, wind and light system i middle g Hydropower station t Real-time outbound flow; Water, wind and light system i middle g -1 The outflow from the hydropower station reached the first g The time of the hydroelectric power station.

[0061] The power output constraints of the hydropower station are as follows:

[0062] In the formula, Indicating a water-wind-solar system i middle g Water consumption rate of hydropower stations; and Representing water, wind, and solar systems respectively i middle g Hydropower station t Maximum and minimum power generation within a time period.

[0063] 2) Bidding curves for multi-regional spot markets of water, wind, and solar power systems An improved supply function equilibrium model was used to construct the bidding curves for each power plant. The bidding curves submitted by each power plant to the regional market operators mainly consist of the winning bid capacity and the generation cost, as shown below:

[0064]

[0065]

[0066] in, Represents water, wind, and light system i In the scene s Down t Always towards d The spot bidding curve submitted by the region, The variable representing the determining coefficient of the linear term of the power bid by the water-wind-solar system i for region d at time t in scenario s is... The variable representing the determining coefficient of the quadratic term of the power bid by the water-wind-solar system i for region d at time t in scenario s is... For the coefficient variable initial value, For the coefficient variable initial value, For the coefficient variable The slope of the change in declared power. For the coefficient variable The slope of the change in declared power.

[0067] 3) Multi-regional spot market optimization and clearing model Regional market operators use the winning bid power of each market participant at each time point as the decision variable, aiming to minimize electricity purchase costs in the spot market optimization and clearing process. Details are as follows:

[0068] In the formula, The total cost of electricity purchase in the regional spot market. The curve representing the spot bidding price submitted by the water-wind-solar system i to region d at time t under scenario s. for s In the scene t Time Zone d The winning bid power of virtual power plant v, where V is the total number of virtual power plants in region d. Let i be the winning bid power of the hydro-wind-solar power generation system i in scenario s at time t in region d. This represents the spot bidding curve submitted by the virtual power plant v located in region d at time t in scenario s.

[0069] The winning bid power for each hydro-wind-solar system in each region must meet the supply and demand balance constraints of the regional spot market and the bid power constraints, as detailed below:

[0070]

[0071]

[0072]

[0073] In the formula, for t Time Zone d The spot market demand power, Water, wind and light system i To the region d The capacity of the transmission lines.

[0074] Step S3: Solve the two-layer model using a distributed iterative solution algorithm.

[0075] 1) Main loop of the distributed iterative solution algorithm This paper decomposes the multi-leader, multi-follower game problem into several subproblems using a distributed iterative algorithm: market bidding and regional market clearing. The approximate global optimum is then solved iteratively. This method not only reduces computational complexity but also effectively utilizes the resources of multiple computing nodes through distributed computing, improving solution efficiency. These subproblems are solved sequentially according to the leader-follower order of the game. First, each market bidding participant, acting as a leader, formulates its bidding strategy. Then, regional market operators, acting as followers, clear the market based on the bidding curves of each market bidding participant. The strategies of leaders and followers are recursively and iteratively adjusted until a Nash equilibrium or the maximum number of iterations is reached. Figure 2 The specific steps are as follows: S31. According to the optimization requirements of the two-level model, input the unit parameters, predicted runoff, predicted power generation, and market demand at each stage for each market bidding participant. S32. Randomly generate initial decision variables for each optimization model in the multi-leader layer, and set the convergence error. And set the iteration count to iter=1, and then start the iteration; S33. Solve the optimization models of the multi-leader layer using the solver to obtain the optimal bidding curve of each market bidding participant in the iter round; S34. Based on the optimal bidding curve obtained in step S33, solve the spot market optimization clearing model for each region in the multi-follower layer, obtain the clearing result of each region in the iter round, and return the result to the multi-leader layer. S35. Starting from the +1th iteration, each market bidding participant in the multi-leader layer updates the bidding curve of the +1th iteration based on the clearing result of the +1st iteration, and inputs the bidding curve into the multi-follower layer to solve for the clearing result of the +1st iteration. S36. Determine whether the solution result of the (iter+1)th round is consistent with the result of the (iter)th round. If they are consistent, the game has reached Nash equilibrium, and the solution result of the (iter+1)th round is the equilibrium solution of the Stackelberg game model. Terminate the iteration. Otherwise, update the iteration count to (iter+2)th and go to step S33 to continue solving until the solution result converges or the maximum number of iterations is reached.

[0076] The convergence error of the iteration is the maximum relative error of the model outputs in two iterations, as shown below:

[0077]

[0078]

[0079] in, Represents the maximum relative error across multiple leadership levels. Represents the maximum relative error across multiple leadership levels. This represents the output of the optimization model l in the iterth round of the multi-leader layer. This represents the output of the optimization model f in the iterth round of the multi-follower layer.

[0080] 2) Quadratic update penalty term of upper-level model To eliminate the tendency for corner solutions in the leader's optimal response problem at the feasible region boundary and to avoid policy oscillations caused by optimal solution jumps, a quadratic update penalty term is introduced into the leader's objective function, making the optimization problem strongly convex. As shown below:

[0081] In the formula, This represents the objective function after adding the penalty term. Represents the original objective function. The decision variable representing the m-th water-wind-solar system bidding strategy optimization model in this round is... represents the decision variable of the previous round's m-th water-wind-solar system bidding strategy optimization model. The smoothing coefficient characterizes the adjustment friction resistance of decision variables during iterative adjustment. M represents the total number of bidding strategy optimization models for the water, wind, and solar power system, equal to the total number of water, wind, and solar power systems, I. Under this penalty term, any allocation that deviates significantly from the previous round's decision will incur additional penalty costs, thus forcing the leader to adopt a smoother strategy adjustment path during iteration. This mechanism effectively enhances the strong convexity of the objective function, making the decision solution more numerically stable and suppressing directional jumps that may be induced by excessive competition at the game theory level.

[0082] 3) Damped Update Method for Lower-Level Model Improvements are made to steps S4 and S5 in the main loop of the distributed iterative solution algorithm (1). After obtaining the clearing results for each region in the (iter+1)th round, instead of directly replacing the results of the previous round with a complete update, a damped update is applied to the clearing results, ensuring that the variables of both the upper and lower level models converge towards the new solution. The specific update form is as follows:

[0083] in, This represents the clearing result of the (iter+1)th round after the update. This represents the clearing result of the iter round. This represents the actual clearing result of round (iter+1). The damping coefficient is used to adjust the degree to which the new solution permeates historical strategies. This damping method effectively suppresses overreactions in strategies caused by competitive behavior, preventing abrupt changes in power and price trajectories during iteration, thus significantly improving the numerical stability of the game-solving process. Theoretically, damped updates can be viewed as a weighted average of the strategy mapping, making it more contractile, thereby enhancing the existence and uniqueness of equilibrium points and accelerating the solution speed of the two-layer model.

[0084] This invention also provides a method for optimized scheduling in a multi-regional spot market involving hydropower, wind power, and solar power. The method includes: constructing a joint optimized scheduling model for hydropower, wind power, and solar power under a power market environment based on the winning bid power obtained from the bidding strategy optimization method for the multi-regional spot market involving hydropower, wind power, and solar power as described above; and obtaining the scheduling plan for each hydropower station in each hydropower, wind power, and solar power system, i.e., the curve of power generation over time. Finally, combining the power generation of the wind and solar power generation units, the scheduling plan for the entire hydropower, wind power, and solar power system is obtained.

[0085] The specific details of the joint optimization scheduling model for water, wind, and solar power are as follows: 1) Objective function: Minimize the water wastage of the water-wind-solar system.

[0086] 2) Hydraulic constraints of cascade hydropower

[0087] The power output constraints of the hydropower station are as follows:

[0088] 3) Contract power balance constraints

[0089] in, Let represent the contract power of the water-wind-solar system i at time t in region d.

[0090] In summary, this invention proposes an optimization method for bidding strategies in multi-regional spot markets involving hydropower, wind power, and solar power based on multi-agent game theory. It employs Latin hypercube sampling and K-means clustering to generate prediction errors for hydropower runoff and wind / solar power output, thus creating a set of uncertain scenarios for runoff and output. A two-layer model for optimizing bidding in multi-regional spot markets involving hydropower, wind power, and solar power based on a multi-leader, multi-follower Stackelberg game is constructed. The upper layer model is an optimization model for bidding strategies of multiple hydropower, wind power, and solar power systems, considering power decomposition curves from medium- and long-term contracts and power purchase agreements. The lower layer model is an optimization clearing model for spot markets in multiple regions. A distributed iterative solution algorithm is used to solve this two-layer model, obtaining the winning bid power in the spot markets for each hydropower, wind power, and solar power system and the clearing price in the spot markets of each region. This algorithm is efficient and ensures the privacy and security of market participants.

[0091] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for optimizing bidding strategies in multi-regional spot markets involving water, wind, and solar power, characterized in that: Includes the following steps: S1. Collect raw power generation data of hydropower station and wind and solar power generation unit at different times. Then, generate typical scenarios of hydropower station and typical scenarios of wind and solar power generation unit by Latin hypercube sampling method and clustering algorithm. By combining them in pairs, obtain S typical scenario sets. The typical scenarios include power generation data change curves over time. S2. Construct a two-layer model for optimizing bidding in the multi-regional spot market based on a multi-leader, multi-follower Stackelberg game involving water, wind, and solar power. The two-layer model includes a bidding strategy optimization model for water, wind, and solar power systems and an optimization clearing model for the multi-regional spot market. The bidding strategy optimization model for water, wind, and solar power systems takes maximizing total revenue as its objective function, while the optimization clearing model for the multi-regional spot market takes the winning bid power as the decision variable and minimizes the electricity purchase cost as its objective. S3. Solve the two-layer model using a distributed iterative solution algorithm to obtain the winning bid power and clearing price of the hydro-wind-solar system in the spot market of various regions.

2. The method for optimizing the bidding strategy of multi-regional spot market involving water, wind, and solar power as described in claim 1, is characterized in that... In step S1, the generation of the typical scenario specifically includes: S11. Divide the interval [0,1] of the cumulative probability distribution function into... N Divide the sample into equal parts, and then obtain the prediction error for each sample n based on the cumulative probability distribution function. Then, based on the original power generation data and prediction error, the power generation data sampling value is calculated to generate the uncertain scenario of hydropower station and wind and solar power generation system; S12. The uncertain scenarios are clustered using the K-means clustering method to generate typical scenarios for hydropower stations and wind and solar power generation units, and the probability of occurrence of each typical scenario is calculated. S13. Use Cartesian products to combine the typical scenarios of the hydropower station and wind and solar power generation units in pairs to obtain S typical scenario sets, and calculate the occurrence probability of each typical scenario set according to the occurrence probability of the typical scenarios.

3. The method for optimizing the bidding strategy of multi-regional spot market participation involving water, wind, and solar power as described in claim 2, is characterized in that... In step S1, the original power generation data of the hydropower station includes the hydropower station runoff variation curve over time, and the original power generation data of the wind and solar power generation system includes the wind and solar power generation power variation curve over time.

4. The method for optimizing the bidding strategy of multi-regional spot market participation involving water, wind, and solar power as described in claim 1, is characterized in that... In step S2, the objective function of the bidding strategy optimization model for the water-wind-solar system is as follows: In the formula, Represents water, wind, and light systems i In the spot market, I represents the total number of water, wind, and solar systems. Indicating a water-wind-solar system i exist s In the scene t Always d The spot market revenue obtained by a region, where D represents the total number of regions and T represents the total duration of spot market power supply. Represents water, wind, and light systems i exist s In the scene t The cost of generating electricity at any given time This represents the probability of scenario s occurring. Represents water, wind, and light systems i In the scene s Down t Always in the region d The winning bid power, Represents water, wind, and light systems i Zhongshui Hydropower Station g exist s In the scene t The planned power generation at any given time, where G represents the total number of hydropower stations in the water-wind-solar system i. This represents the clearing electricity price in region d at time t under scenario s. The power generation cost coefficient of hydropower station g in the water-wind-solar system i. This represents the power generation of wind and solar power system j within the water-wind-solar system i at time s. The power generation cost coefficient represents the power generation cost of wind and solar power system j within the water-wind-solar system i.

5. The method for optimizing the bidding strategy of multi-regional spot market participation involving water, wind, and solar power as described in claim 4, is characterized in that... The bidding strategy optimization model for the hydro-wind-solar system includes bidding power constraints, cascade hydropower station operation constraints, and hydropower station output constraints; the bidding power constraints are as follows: In the formula, Represents water, wind, and light systems i exist s In the scene t Always towards d The bidding power submitted by the region; The operational constraints of the cascade hydropower stations are as follows: In the formula, , and Representing water, wind, and solar systems respectively i middle g hydroelectric power station t Water level at the end of the time period and its upper and lower limits; Indicating a water-wind-solar system i middle g The reservoir water level-capacity curve function of a hydroelectric power station; and These represent the initial water level constraint and the final water level constraint, respectively. , and Representing water, wind, and solar systems respectively i middle g hydroelectric power station t The discharge flow rate and its upper and lower limits for each time period; , and Representing water, wind, and solar systems respectively i middle g Hydropower station t The power generation flow rate at any given time and its upper and lower limits; , Representing water, wind, and solar systems respectively i middle g Hydropower station t Storage capacity at time t1 and time t-1; For time intervals; R i,g,s,t For water, wind and light system i middle g Hydropower station t Inbound traffic at any given time; I i,g,s,t For water, wind and light system i middle g Hydropower station t Real-time outbound flow; For water, wind and light system i middle g -1 The outflow from the hydropower station reached the first g The time for each hydroelectric power station; The power output constraints of the hydropower station are as follows: In the formula, Indicating a water-wind-solar system i middle g Water consumption rate of hydropower stations; and Representing water, wind, and solar systems respectively i middle g Hydropower station t Maximum and minimum power generation within the time period; The bidding curve consists of the bidding power and the generation cost, as shown below: in, Represents water, wind, and light systems i In the scene s Down t Always towards d The spot bidding curve submitted by the region, The variable representing the determining coefficient of the linear term of the power bid by the water-wind-solar system i for region d at time t in scenario s is denoted by . The variable representing the determining coefficient of the quadratic term of the power bid by the water-wind-solar system i for region d at time t in scenario s is... For the coefficient variable initial value, For the coefficient variable initial value, For the coefficient variable The slope of the change in declared power. For the coefficient variable The slope of the change in declared power.

6. The method for optimizing the bidding strategy of multi-regional spot market participation involving water, wind, and solar power as described in claim 5, is characterized in that... The objective function of the multi-regional spot market optimization clearing model is shown below: In the formula, The total cost of electricity purchase in the regional spot market. The curve representing the spot bidding price submitted by the water-wind-solar system i to region d at time t under scenario s. for s In the scene t Time Zone d The winning bid power of virtual power plant v, where V is the total number of virtual power plants in region d. Let i be the winning bid power of the hydro-wind-solar power generation system i in scenario s at time t in region d. The spot bidding curve for virtual power plant v located in region d at time t in scenario s; And set regional spot market supply and demand balance constraints and bidding power constraints, as shown below: In the formula, for t Time Zone d The spot market demand power, For water, wind and light system i To the region d The capacity of the transmission lines.

7. The method for optimizing the bidding strategy of multi-regional spot market participation involving water, wind, and solar power as described in claim 6, is characterized in that... In step S3, the iterative solution of the two-layer model includes the following steps: S31. According to the optimization requirements of the two-level model, input the unit parameters, predicted runoff, predicted power generation, and market demand at each stage for each market bidding participant. S32. Randomly generate initial decision variables for each optimization model in the multi-leader layer, and set the convergence error. And set the iteration count to iter=1, and then start the iteration; S33. Solve the optimization models of the multi-leader layer using the solver to obtain the optimal bidding curve of each market bidding participant in the iter round; S34. Based on the optimal bidding curve obtained in step S33, solve the spot market optimization clearing model for each region in the multi-follower layer, obtain the clearing result of each region in the iter round, and return the result to the multi-leader layer. S35. Starting from the +1th iteration, each market bidding participant in the multi-leader layer updates the bidding curve of the +1th iteration based on the clearing result of the +1st iteration, and inputs the bidding curve into the multi-follower layer to solve for the clearing result of the +1st iteration. S36. Determine whether the solution result of the (iter+1)th round is consistent with the result of the (iter)th round. If they are consistent, the game has reached Nash equilibrium, and the solution result of the (iter+1)th round is the equilibrium solution of the Stackelberg game model. Terminate the iteration. Otherwise, update the iteration count to (iter+2)th and go to step S33 to continue solving until the solution result converges or the maximum number of iterations is reached.

8. The method for optimizing the bidding strategy of multi-regional spot market participation involving water, wind, and solar power as described in claim 7, is characterized in that... The iterative solution is further optimized using a quadratic update penalty term and a damped update method; the quadratic update penalty term is shown below: In the formula, This represents the objective function after adding the penalty term. Represents the original objective function. The decision variable representing the m-th water-wind-solar system bidding strategy optimization model in this round is... represents the decision variable of the previous round's m-th water-wind-solar system bidding strategy optimization model. The smoothing coefficient is used to characterize the adjustment friction resistance of decision variables during the iterative adjustment process, and M represents the total number of bidding strategy optimization models for water, wind and solar systems. The damping update method includes: after obtaining the clearing results of each region in the (iter+1)th round, implementing damping update on the clearing results to ensure that the variables of the upper and lower level models converge towards the new solution. The specific update form is as follows: in, This represents the clearing result of the (iter+1)th round after the update. This represents the clearing result of the iter round. This represents the actual clearing result of round (iter+1). This represents the damping coefficient, used to adjust the degree to which a new solution permeates historical strategies.

9. A method for optimizing the scheduling of multi-regional spot markets involving water, wind, and solar power, characterized in that: include: Based on the winning bid power obtained by the multi-regional spot market bidding strategy optimization method for joint participation of hydropower, wind power and solar power as described in any one of claims 1-8, a joint optimization scheduling model for hydropower, wind power and solar power under the power market environment is constructed to obtain the scheduling plan for each hydropower station in each hydropower, wind power and solar power system.

10. The multi-regional spot market optimization scheduling method involving water, wind, and solar power as described in claim 9, is characterized in that... The objective function of the joint optimization scheduling model for water, wind, and solar power is shown below: The hydraulic constraints of the cascade hydropower project are shown below: The power output constraints of the hydropower station are as follows: The contract power balance constraints are as follows: in, Let represent the contract power of the water-wind-solar system i at time t in region d.

Images