Unit output and offer collaborative optimization method and system for power spot market

Abstract

The application discloses a unit output and offer collaborative optimization method and system for a power spot market, and belongs to the technical field of power spot market optimization decision-making. The method comprises the following steps: obtaining operation data and market data of a thermal power unit, and constructing an input data matrix containing a time sequence based on the operation data and the market data; establishing a unified output-offer collaborative optimization objective function based on the input data matrix, wherein the objective function comprises a unit operation cost item, a market income item and a risk control item, and weight coefficients are assigned to each item; embedding physical constraints of unit operation and market transaction rule constraints as strong constraint conditions into the output-offer collaborative optimization objective function; iteratively solving the output-offer collaborative optimization objective function embedded with the strong constraint conditions by using a multi-stage energy disturbance collaborative mechanism; and outputting a collaborative optimization result containing an optimal offer curve and a contract allocation ratio when a convergence condition is met.

Description

Technical Field

[0001] This invention belongs to the field of power spot market optimization decision-making technology, and relates to a method and system for coordinating optimization of unit output and bidding in the power spot market. Background Technology

[0002] With the continued deepening of power market reform and the full implementation of the electricity spot market, the traditional planned power generation and fixed-price operation model of thermal power units has been completely changed. Thermal power enterprises have transformed from traditional production-oriented entities to market-oriented operators that emphasize both production and market operation. Under the electricity spot market trading mechanism, the daily operation of thermal power units faces constraints and challenges from a dual core management system: On the one hand, unit production and operation must strictly adhere to hard physical conditions such as grid dispatching regulations, unit physical operating limits, and grid safety and stability constraints, while also taking into account the refined management of various production costs such as coal consumption, operation and maintenance, and start-up and shutdown, to ensure the safe, stable, and economical operation of the units; on the other hand, electricity spot market prices fluctuate in real time, the trading pace is fast, and the bidding rules are complex. The day-ahead and intraday real-time bidding strategies of the units directly determine the enterprise's power generation revenue and market competitiveness. The scientific and accurate nature of bidding decisions plays a decisive role in the profitability of thermal power enterprises. Currently, the operation optimization system of thermal power enterprises for the spot market is mainly built independently around two modules: unit production and operation optimization and market bidding decision-making. Conventional technical solutions in the industry are mostly designed to optimize a single aspect. Among them, production operation optimization technology focuses only on optimizing production-side indicators such as unit load allocation, parameter adjustment, and energy consumption control, with the core objective of meeting the physical operation constraints of the units and reducing the cost of single production; while market bidding decision technology mainly relies on historical electricity price data and load forecast data to conduct bidding strategy deduction, with the core objective of maximizing the profit of a single transaction.

[0003] However, in the actual operation of the electricity spot market, production operation and market bidding are highly coupled and mutually restrictive. Existing single-dimensional and fragmented optimization technologies are gradually revealing many inherent defects and cannot adapt to the optimal operation needs of thermal power plants across the entire market scenario. The specific technical problems are as follows: First, the production decision-making and marketing bidding decision-making systems are fragmented, lacking an integrated and unified optimization framework. Existing technologies separate unit production operation optimization and market bidding decisions into two independent modules. Production-side optimization does not match market price fluctuation patterns, and bidding-side decisions do not fully consider the real-time physical constraints and cost characteristics of unit operation. This leads to a disconnect between production plans and market pricing strategies, easily resulting in contradictions such as "the optimal production plan cannot meet the market bidding revenue requirements, and the optimal bidding strategy exceeds the actual operating capacity of the unit," failing to achieve optimal synergy across the entire production and marketing process. Second, multiple optimization objectives inherently conflict, making effective trade-offs and adaptation difficult. Market-oriented operation of thermal power enterprises requires simultaneously considering multiple objectives such as safe and stable unit operation, minimum energy consumption costs, maximum market bidding revenue, and minimum unit losses. However, existing optimization technologies often adopt a single-objective weighted simplified optimization model, which cannot accurately quantify the coupling constraints and conflict relationships between objectives. It is difficult to dynamically trade off and coordinate multi-dimensional objectives, often resulting in optimal single indicators but poor overall operational efficiency, failing to meet the multi-objective optimization needs of complex spot market scenarios. Third, traditional optimization algorithms have weak global search capabilities and are prone to getting trapped in local optima, leading to poor stability of enterprise revenue. Existing conventional optimization algorithms mostly operate on a local iterative optimization model with a limited convergence range. In the complex optimization scenarios of the electricity spot market, characterized by multiple constraints, variables, and strong coupling, they are highly susceptible to converging to local optima, unable to traverse all feasible decision schemes, and struggling to lock in the globally optimal production bidding coordination strategy. This results in large fluctuations in unit power generation revenue, low overall revenue, and an inability to guarantee the stability of enterprise market-based operation revenue. Fourth, they lack robust handling capabilities against market uncertainties, leading to poor adaptability of optimization schemes. Electricity spot market prices are affected by multiple factors such as electricity load, renewable energy output, grid supply and demand, and inter-provincial transactions, exhibiting strong randomness, volatility, and uncertainty. Existing optimization techniques mostly rely on deterministic forecast data for static optimization, failing to consider the impact of uncertain risks such as electricity price fluctuations and load deviations on production and bidding decisions. The optimization schemes have weak anti-interference capabilities and are prone to failure when market conditions fluctuate, further exacerbating the operational risks and revenue uncertainties of thermal power enterprises.

[0004] In summary, existing spot market operation optimization technologies for thermal power enterprises suffer from core defects such as fragmented decision-making, failure of multi-objective trade-offs, local optima in optimization results, and poor resistance to uncertainty interference. These defects severely restrict the safe operation capabilities and profitability of thermal power enterprises in the spot market. Summary of the Invention

[0005] The purpose of this invention is to solve the technical problems of fragmented decision-making, failure of multi-objective trade-offs, local optima in optimization results, and poor resistance to uncertainty in the existing thermal power spot market operation optimization technology, and to provide a method and system for coordinated optimization of unit output and bidding for the electricity spot market.

[0006] To achieve the above objectives, the present invention employs the following technical solution: In a first aspect, this invention discloses a method for coordinating and optimizing generator output and bidding prices in the electricity spot market, comprising: Obtain operating data and market data of thermal power units, and construct an input data matrix containing time series data based on the operating data and market data; Based on the input data matrix, a unified output-price collaborative optimization objective function is established. The objective function includes unit operating cost, market revenue, and risk control items, and weight coefficients are assigned to each item. The physical constraints of unit operation and the constraints of market trading rules are used as strong constraints and embedded into the output-price collaborative optimization objective function; A multi-stage energy perturbation coordination mechanism is used to iteratively solve the output-bid coordination optimization objective function with embedded strong constraints; when the convergence condition is met, the output includes the coordination optimization result containing the optimal bid curve and the contract allocation ratio.

[0007] Further improvements are made in the following aspects: The multi-stage energy perturbation coordination mechanism includes: Based on individual energy levels and local environment density, an adaptive global exploration or local exploitation strategy is selected to update candidate solutions; When the optimization process is detected to be trapped in a local optimum, a negative curvature escape operation is performed to escape the local region.

[0008] The acquisition of operating data and market data of thermal power units, and the construction of an input data matrix containing time series data based on the operating data and market data specifically include: Basic data of thermal power units within the dispatch cycle are collected, including historical output sequences, spot market price sequences, fuel cost data, and meteorological influencing factors. This data is aligned according to time granularity, and the dispatch cycle is divided into [number] periods. Construct the input matrix at discrete time points:

[0009] in, For the input data matrix; The number of feature dimensions; This represents the length of the time series.

[0010] The unified output-price collaborative optimization objective function established based on the input data matrix includes: Define the output-quote collaborative optimization objective function as follows:

[0011] in, To optimize the objective function value for output-quote collaboration, - The target weight coefficient; The sub-objectives are defined as follows: Fuel cost item:

[0012] in, This is the unit fuel cost coefficient for electricity generation. For a moment The unit output; Start-up and shutdown costs:

[0013] in, This refers to the start-up and shutdown cost coefficient. This is a unit status variable, where 0 indicates shutdown and 1 indicates operation; Climbing stability:

[0014] in, This is the climbing penalty coefficient; Market returns:

[0015] in, For a moment The quote; Deviation assessment:

[0016] in, For planned or settled electricity consumption; This is the deviation penalty coefficient; risk:

[0017] in, Let V be the price volatility variance.

[0018] The inclusion of physical constraints on unit operation and market trading rules as strong constraints in the output-price collaborative optimization objective function includes: The physical constraints and market trading rules governing the operation of the generating units include: Output boundary constraints:

[0019] Climbing constraints:

[0020] Quotation constraints:

[0021] Contract allocation constraint: Assume the contract allocation ratio is... ,satisfy

[0022] in, These are the unit's minimum and maximum outputs; It is the maximum climbing rate; These are the upper and lower limits of market prices.

[0023] The iterative solution of the output-bid co-optimization objective function with embedded strong constraints using a multi-stage energy perturbation cooperative mechanism includes: Candidate solution initialization: Let the population size be Each candidate solution is represented as:

[0024] in, For the price quote sequence vector, Assign a vector to the contract; Energy-driven search mechanism: Define individual energy:

[0025] in, For the first The individual in the first Energy of the next iteration; This is the distance consumption coefficient; Update the distance to the location; Return enhancement factor; The improvement amount of the objective function; Environmental density regulation mechanism: Define local density as The adjustment factor is:

[0026] Used to control the search step size; Adaptive position update: when Perform a global update at the specified time:

[0027] Otherwise, perform a partial update: in, This is the current optimal solution; For random neighbor solutions; This is the iteration step size factor; Negative curvature escape mechanism: When a negative local curvature is detected:

[0028] in, The direction is random; The escape amplitude is used to avoid getting trapped in local optima. Random exploration and smoothing mechanism: Introduce random perturbations:

[0029] And perform exponential smoothing:

[0030] in, Step size; The disturbance intensity; It is random noise; This is the smoothing coefficient.

[0031] When the convergence condition is met, the output includes the co-optimization result containing the optimal bid curve and the contract allocation ratio, including: Once the iteration converges, the optimal decision result is output:

[0032] in, This represents the optimal pricing curve. To achieve the optimal contract allocation ratio; The optimal decision result is directly used for market application and production scheduling; After actual operation, obtain the actual clearing results, deviation assessment results, and profit realization status, and update the target weights. Risk parameters and optimized initial population are used to form a continuous optimization iteration.

[0033] Secondly, this invention discloses a system for coordinating and optimizing generator output and bidding prices in the electricity spot market, comprising: The data preparation module is used to acquire the operating data and market data of thermal power units, and to construct an input data matrix containing time series data based on the operating data and market data. The multi-objective modeling module is used to establish a unified output-price collaborative optimization objective function based on the input data matrix. The objective function includes unit operating cost, market revenue, and risk control, and assigns weight coefficients to each objective. The constraint processing module is used to embed the physical constraints of unit operation and market trading rule constraints as strong constraints into the output-price collaborative optimization objective function; The collaborative optimization and decision output module is used to iteratively solve the output-bid collaborative optimization objective function with embedded strong constraints using a multi-stage energy perturbation collaborative mechanism; when the convergence condition is met, it outputs the collaborative optimization result containing the optimal bid curve and the contract allocation ratio.

[0034] Thirdly, the present invention discloses an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-mentioned method for coordinating and optimizing unit output and bidding for the electricity spot market.

[0035] Fourthly, the present invention discloses a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-mentioned method for coordinating and optimizing unit output and pricing for the electricity spot market.

[0036] Compared with the prior art, the present invention has the following beneficial effects: This invention discloses a method for co-optimizing generator output and pricing in the electricity spot market. First, it achieves deep collaboration between production and marketing decisions. By establishing a unified output-pricing co-optimization objective function, this invention incorporates generator operating costs, market revenue, and risk control into the same optimization framework, avoiding the drawbacks of traditional methods where production planning and pricing strategies are formulated separately and are fragmented. This method enables output adjustments and pricing changes to respond synchronously to market signals, forming a closed-loop linkage between the production and marketing sides, thereby improving the consistency and economy of overall decision-making from the source. Second, it ensures the engineering feasibility of the optimization results while meeting strict constraints. This invention directly embeds the physical constraints of generator operation (such as output boundaries and ramp rates) and market trading rule constraints (such as price limits and contract allocation ratios) as strong constraints into the optimization process, rather than using soft constraints such as penalty functions. This design ensures that every candidate solution generated by iterative solving naturally meets the requirements of safe generator operation and market compliance, avoiding the risk of optimization results deviating from actual operating capacity or illegal pricing in traditional methods, and greatly improving the engineering feasibility of the decision results. Third, it effectively balances global search capability with local convergence accuracy, avoiding getting trapped in local optima. This invention employs a multi-stage energy perturbation collaborative mechanism for iterative solution. This mechanism adaptively switches between the global exploration stage and the local development stage by simulating the dynamic changes in individual energy levels and environmental density. Compared to the fixed search strategy in traditional heuristic algorithms, this method can fully traverse the solution space in the early stages of optimization to discover high-potential regions, and automatically shrink the search range in the later stages of optimization to achieve fine convergence. Simultaneously, this mechanism possesses an inherent ability to escape local optima; when search stagnation is detected, it can proactively introduce perturbations, thereby enabling the algorithm to continuously evolve towards a better direction, significantly improving the stability and profitability of the optimization results. Fourth, it exhibits good robustness against uncertainties such as market price fluctuations. This invention explicitly introduces risk control terms (e.g., price volatility variance) into the objective function and uses the price curve and contract allocation ratio as collaborative decision variables. This allows the optimization process to not only pursue profit maximization but also consider profit stability, automatically adjusting the price strategy and contract allocation ratio according to the degree of market volatility, thereby achieving more robust overall benefits in a volatile spot market environment. Compared to traditional methods that only focus on expected returns, this invention is closer to the decision-making needs of thermal power companies in actual operation, which require both increasing revenue and mitigating risks. Fifth, it possesses good scalability and adaptability. The objective function framework and multi-stage energy disturbance coordination mechanism constructed in this invention do not depend on specific unit parameters or market rule details. By adjusting the weight coefficients of each objective item or replacing the constraints, it can be adapted to thermal power units of different capacity levels, different spot market trading rules (such as segmented pricing and nodal pricing), and complex scenarios such as the connection between medium- and long-term markets and the spot market.Furthermore, this method can be easily extended to multi-unit joint optimization or regional power combination decision-making, possessing high engineering reuse value. In summary, this invention systematically solves the technical challenges of production and marketing synergy optimization in the context of thermal power enterprises participating in the electricity spot market, achieving superior progress compared to existing technologies in terms of decision quality, constraint satisfaction capability, global optimization performance, risk robustness, and engineering applicability.

[0037] This invention discloses a collaborative optimization system for generating unit output and pricing in the electricity spot market. First, its modular architecture achieves clear decoupling and efficient integration of the production and marketing collaborative optimization processes. The system is divided into modules according to a logical chain of data preparation, multi-objective modeling, constraint handling, collaborative optimization, and decision output. Each module has clearly defined functional boundaries and standardized data interfaces. The data preparation module is responsible for collecting and aligning raw data, providing a unified input benchmark for subsequent modeling. The multi-objective modeling module and the constraint handling module construct the mathematical expression of the optimization problem in parallel, allowing for independent addition, deletion, and adjustment of objective functions and constraints without affecting the core logic of other modules. The collaborative optimization and decision output module serves as the solution engine and is loosely coupled with the front-end modeling module. This architectural design facilitates rapid deployment and maintenance of the system within existing information platforms in thermal power plants and supports module-level replacement or upgrades for different market rules or generating unit characteristics, greatly improving the system's engineering adaptability and scalability. Second, the tight coupling design between the constraint handling module and the multi-objective modeling module ensures the physical feasibility and market compliance of the optimization results. The constraint processing module in this system does not simply filter or penalize solutions in the later stages of optimization. Instead, it directly embeds the physical constraints of unit operation (such as output boundaries and ramp rates) and market trading rule constraints (such as bid limits and contract allocation ratios) as strong constraints into the solution space of the output-bid co-optimization objective function. During the iterative solution process, the co-optimization and decision output module always generates candidate solutions only within the feasible region. This mechanism ensures that the optimal bid curve and contract allocation ratio output by the system meet the requirements for safe unit operation and electricity market declaration rules without any post-processing corrections, eliminating the risk of "feasible but unusable" optimization results in traditional systems and significantly improving the engineering executability of the decision results. Third, the multi-stage energy perturbation coordination mechanism built into the co-optimization and decision output module endows the system with powerful global optimization capabilities and local convergence stability. During the solution process, this module dynamically simulates individual energy levels and environmental density, adaptively switching between global exploration and local development, and automatically executing escape operations when search stagnation is detected. Compared to existing systems employing conventional heuristic algorithms (such as genetic algorithms and particle swarm optimization), this system significantly reduces the probability of getting trapped in local optima when facing high-dimensional, nonlinear, and multi-peak output-bid co-optimization problems, thus achieving better overall returns with the same computational resources. Simultaneously, this module integrates convergence determination and decision output into the same process. Upon meeting the convergence conditions, it directly outputs the bid curve and contract allocation ratio, avoiding additional manual judgment and improving decision response speed. Fourth, the system possesses inherent robustness in handling market uncertainty. The multi-objective modeling module explicitly incorporates risk control terms (such as price volatility variance) when constructing the objective function and weights them with the market return term.This means that when the system performs collaborative optimization, it does not simply pursue maximum returns, but simultaneously weighs the volatility risk of returns. Driven by this risk-return joint objective, the collaborative optimization and decision output module can automatically adjust the shape of the price curve and the contract allocation ratio according to different market volatility levels, for example, proactively reducing risk exposure during periods of high volatility. This endogenous risk handling mechanism enables the system's output decision results to maintain relatively robust overall benefits under different market environments, more closely aligning with the decision-making preferences of thermal power enterprises in actual operation, which balance increasing revenue and mitigating risks. Fifth, the seamless integration between the data preparation module and the front-end module provides high-quality data support for the system. The data preparation module aligns historical output sequences, spot price sequences, fuel cost data, and meteorological influencing factors at the time granularity according to the scheduling cycle (e.g., 96 points), constructing a structured input data matrix. This matrix not only provides a unified time benchmark for the multi-objective modeling module but also facilitates batch vectorized calculations by subsequent modules. Compared with existing systems with scattered and inconsistent data formats, this invention, through a standardized data preparation process, reduces optimization deviations caused by data quality issues, improving the stability of system operation and the reproducibility of results. In summary, the unit output and bidding collaborative optimization system proposed in this invention for the electricity spot market achieves superior technical results compared to existing systems in terms of decision quality, constraint satisfaction capability, global optimization performance, risk robustness, and ease of engineering deployment through reasonable module division and functional collaboration. It provides thermal power companies with a reliable and efficient decision support tool for participating in the electricity spot market. Attached Figure Description

[0038] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0039] Figure 1 This is a flowchart of a method for coordinating and optimizing generator output and bidding in the electricity spot market, as described in an embodiment of the present invention. Figure 2 This is a block diagram of a generator output and bidding coordination optimization system for the electricity spot market, as described in an embodiment of the present invention. Detailed Implementation

[0040] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0041] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0042] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0043] The present invention will now be described in further detail with reference to the accompanying drawings: See Figure 1 This invention discloses a method for coordinating and optimizing generator output and bidding prices in the electricity spot market, comprising: S1, acquire the operating data and market data of the thermal power unit, and construct an input data matrix containing time series based on the operating data and market data; S2, Based on the input data matrix, establish a unified output-price collaborative optimization objective function. The objective function includes unit operating cost, market revenue, and risk control items, and assign weight coefficients to each item. S3, the physical constraints of unit operation and market trading rules are used as strong constraints and embedded into the output-price collaborative optimization objective function; S4 employs a multi-stage energy perturbation collaborative mechanism to iteratively solve the output-bid collaborative optimization objective function with embedded strong constraints; when the convergence condition is met, the output includes the collaborative optimization result containing the optimal bid curve and the contract allocation ratio.

[0044] This invention discloses a method for co-optimizing generator output and pricing in the electricity spot market. First, it achieves deep collaboration between production and marketing decisions. By establishing a unified output-pricing co-optimization objective function, this invention incorporates generator operating costs, market revenue, and risk control into the same optimization framework, avoiding the drawbacks of traditional methods where production planning and pricing strategies are formulated separately and are fragmented. This method enables output adjustments and pricing changes to respond synchronously to market signals, forming a closed-loop linkage between the production and marketing sides, thereby improving the consistency and economy of overall decision-making from the source. Second, it ensures the engineering feasibility of the optimization results while meeting strict constraints. This invention directly embeds the physical constraints of generator operation (such as output boundaries and ramp rates) and market trading rule constraints (such as price limits and contract allocation ratios) as strong constraints into the optimization process, rather than using soft constraints such as penalty functions. This design ensures that every candidate solution generated by iterative solving naturally meets the requirements of safe generator operation and market compliance, avoiding the risk of optimization results deviating from actual operating capacity or illegal pricing in traditional methods, and greatly improving the engineering feasibility of the decision results. Third, it effectively balances global search capability with local convergence accuracy, avoiding getting trapped in local optima. This invention employs a multi-stage energy perturbation collaborative mechanism for iterative solution. This mechanism adaptively switches between the global exploration stage and the local development stage by simulating the dynamic changes in individual energy levels and environmental density. Compared to the fixed search strategy in traditional heuristic algorithms, this method can fully traverse the solution space in the early stages of optimization to discover high-potential regions, and automatically shrink the search range in the later stages of optimization to achieve fine convergence. Simultaneously, this mechanism possesses an inherent ability to escape local optima; when search stagnation is detected, it can proactively introduce perturbations, thereby enabling the algorithm to continuously evolve towards a better direction, significantly improving the stability and profitability of the optimization results. Fourth, it exhibits good robustness against uncertainties such as market price fluctuations. This invention explicitly introduces risk control terms (e.g., price volatility variance) into the objective function and uses the price curve and contract allocation ratio as collaborative decision variables. This allows the optimization process to not only pursue profit maximization but also consider profit stability, automatically adjusting the price strategy and contract allocation ratio according to the degree of market volatility, thereby achieving more robust overall benefits in a volatile spot market environment. Compared to traditional methods that only focus on expected returns, this invention is closer to the decision-making needs of thermal power companies in actual operation, which require both increasing revenue and mitigating risks. Fifth, it possesses good scalability and adaptability. The objective function framework and multi-stage energy disturbance coordination mechanism constructed in this invention do not depend on specific unit parameters or market rule details. By adjusting the weight coefficients of each objective item or replacing the constraints, it can be adapted to thermal power units of different capacity levels, different spot market trading rules (such as segmented pricing and nodal pricing), and complex scenarios such as the connection between medium- and long-term markets and the spot market.Furthermore, this method can be easily extended to multi-unit joint optimization or regional power combination decision-making, possessing high engineering reuse value. In summary, this invention systematically solves the technical challenges of production and marketing synergy optimization in the context of thermal power enterprises participating in the electricity spot market, achieving superior progress compared to existing technologies in terms of decision quality, constraint satisfaction capability, global optimization performance, risk robustness, and engineering applicability.

[0045] The present invention will be described in detail below with reference to specific embodiments: This embodiment provides a method for coordinating and optimizing generator output and bidding prices in the electricity spot market. The specific steps are as follows: Step 1: Data Input Basic data of thermal power units within the dispatch cycle are collected, including historical output sequences, spot market price sequences, fuel cost data (such as coal prices), and meteorological influencing factors (temperature, humidity, etc.). This data is aligned according to time granularity, and the dispatch cycle is divided into [number missing]. Construct the input matrix using discrete time points (e.g., 96 points):

[0046] in, For the input data matrix; The number of feature dimensions; This represents the length of the time series.

[0047] Step 2: Multi-objective modeling Construct a unified production-marketing collaborative optimization objective function to achieve quantitative expression and weighted fusion of multiple objectives.

[0048] Define the comprehensive optimization objective function as follows:

[0049] in, To optimize the objective function value for output-quote collaboration, - The target weight coefficient; The sub-objectives are defined as follows: Fuel cost item:

[0050] in, This is the unit fuel cost coefficient for electricity generation. For a moment The unit output; Start-up and shutdown costs:

[0051] in, This refers to the start-up and shutdown cost coefficient. This is a unit status variable, where 0 indicates shutdown and 1 indicates operation; Climbing stability:

[0052] in, This is the climbing penalty coefficient; Market returns:

[0053] in, For a moment The quote; Deviation assessment:

[0054] in, For planned or settled electricity consumption; This is the deviation penalty coefficient; risk:

[0055] in, Let V be the price volatility variance.

[0056] Step 3: Constraint Handling To ensure that the optimization results meet the requirements of unit operation and market rules, a strong constraint approach is adopted to embed the optimization process.

[0057] The physical constraints and market trading rules governing the operation of the generating units include: Output boundary constraints:

[0058] Climbing constraints:

[0059] Quotation constraints:

[0060] Contract allocation constraint: Assume the contract allocation ratio is... ,satisfy

[0061] in, These are the unit's minimum and maximum outputs; It is the maximum climbing rate; These are the upper and lower limits of market prices.

[0062] By constraining projection or feasible region correction, we ensure that all candidate solutions satisfy physical and market rules.

[0063] Step 4: Collaborative Optimization A collaborative optimization module is constructed to achieve a dynamic balance between global search and local optimization through a multi-stage energy perturbation collaborative mechanism.

[0064] (1) Initialization of candidate solutions Let the population size be Each candidate solution is represented as:

[0065] in, For the price quote sequence vector, Assign a vector to the contract.

[0066] (2) Energy-driven search mechanism Define individual energy:

[0067] in, For the first The individual in the first Energy of the next iteration; This is the distance consumption coefficient; Update the distance to the location; Return enhancement factor; The improvement amount of the objective function; Environmental density regulation mechanism: Define local density as The adjustment factor is:

[0068] Used to control the search step size; Adaptive position update: when Perform a global update at the specified time:

[0069] Otherwise, perform a partial update: in, This is the current optimal solution; For random neighbor solutions; This is the iteration step size factor; Negative curvature escape mechanism: When a negative local curvature is detected:

[0070] in, The direction is random; This is the escape amplitude, used to avoid getting trapped in local optima.

[0071] Random exploration and smoothing mechanism: Introduce random perturbations:

[0072] And perform exponential smoothing:

[0073] in, Step size; The disturbance intensity; It is random noise; This is the smoothing coefficient.

[0074] Step 5: Decision Output Once the iteration converges, the optimal decision result is output:

[0075] in: Optimal pricing curve; Optimal contract allocation ratio; The results are directly used for market declaration and production scheduling.

[0076] Step Six: Feedback Iteration After actual operation, obtain the actual clearing results, deviation assessment results, and profit realization status, and update the target weights. Risk parameters are used to optimize the initial population. This leads to continuous optimization iterations and improved model adaptability.

[0077] Example 1 Let's take a thermal power plant in a certain province as an example for illustration.

[0078] Taking a thermal power plant in a certain province as an example: Unit scale: 2×600MW, time scale: 96 points (15 minutes), data acquired include: historical output, spot price, coal price, and meteorological data.

[0079] Optimization process: (1) Initialize 100 candidate strategies; (2) Run a multi-stage energy perturbation optimization algorithm; (3) Iterate 200 times to obtain the optimal strategy; Optimization results show that, under the same data conditions, the method described in this invention significantly improves the returns compared to traditional optimization methods.

[0080] Furthermore, the following are some alternative implementation methods: (1) In the multi-objective modeling module, the construction method of the comprehensive optimization objective function is not limited to one. Its multi-objective fusion form can be replaced by weighted summation, hierarchical optimization or Pareto optimal set construction according to application requirements. The weight coefficients of each objective can be set as fixed parameters or adaptively adjusted according to changes in the market environment to adapt to different market volatility levels and differences in return risk preferences.

[0081] (2) In the process of multi-objective modeling, the specific expression of each sub-objective function is not limited to a single form. For example, functions such as fuel cost, start-up and shutdown cost, ramp-up penalty, market revenue and risk measurement can be replaced with linear function, piecewise function or nonlinear function, or a parameterized model based on historical data fitting can be introduced to enhance the ability to characterize different unit characteristics and market mechanisms.

[0082] (3) In the constraint processing module, the strong constraint embedding method is not limited to one. Its constraint processing mechanism can be replaced by penalty function form, feasible domain mapping method or decomposition constraint processing method. The constraint conditions can be extended or adjusted according to the unit type, market rules or dispatch requirements to adapt to different power market operation mechanisms.

[0083] (4) In the constraint modeling stage, the specific forms of the unit operation constraints, quotation constraints and contract allocation constraints are not limited to one. For example, minimum start-up and shutdown time constraints, start-up and shutdown sequence constraints or auxiliary service constraints can be added to enhance the engineering executability and scheduling consistency of the optimization results.

[0084] (5) In the collaborative optimization module, the specific implementation of the multi-stage energy perturbation collaborative mechanism is not limited to one. Its group search strategy can be replaced by other heuristic optimization methods or deterministic optimization methods, including but not limited to optimization mechanisms based on group evolution, path search or gradient update, to adapt to optimization problems of different scales and complexities.

[0085] (6) In the collaborative optimization process, the specific expression of the individual energy-driven update mechanism, the environmental density adjustment mechanism and the random perturbation mechanism is not limited to a single form. The parameter setting method can be replaced by fixed parameters, adaptive parameters or parameter update method that is dynamically adjusted based on historical optimization effect feedback, so as to improve the stability and convergence efficiency of the optimization process.

[0086] (7) In the optimization iteration stage, the stage division method of the search process is not limited to one. The optimization process can be replaced by a multi-stage hierarchical search structure or a continuous adaptive search structure, such as dividing it into a global exploration stage and a local convergence stage, or achieving a balance between exploration and development through a single dynamic mechanism to adapt to different problem scales and convergence speed requirements.

[0087] (8) In the decision output module, the output form of the optimization result is not limited to a single optimal strategy or a set of multiple strategies. It can also be classified and output according to the risk level or return level, such as forming a pricing strategy or contract allocation scheme with different risk levels to meet the diverse decision-making needs in actual operation.

[0088] (9) In the decision output stage, the expression form of the quotation curve and the contract allocation ratio is not limited to a single form. It can be replaced by a segmented quotation form, an interval quotation form or a continuous quotation expression form based on function fitting, in order to adapt to the quotation rules of different electricity markets.

[0089] (10) In the feedback iteration module, the feedback update mechanism is not limited to one. Its parameter update method can be replaced by a statistical correction method based on historical running results, an online learning update method, or an experience reuse method based on the strategy library, so as to gradually improve the adaptability and stability of the optimization strategy.

[0090] (11) At the system implementation level, the deployment method of each functional module is not limited to one. It can be implemented by centralized deployment, distributed deployment or microservice architecture. The modules interact with each other through data interface or message mechanism to adapt to different scale power systems and computing resource environments.

[0091] (12) At the application scenario level, the method and system described in this invention are not only applicable to the production and marketing collaborative optimization of a single thermal power unit, but can also be extended to scenarios such as multi-unit joint optimization, regional power combination optimization, electricity spot and medium- and long-term market collaborative decision-making, and integrated energy system optimization.

[0092] The key technology of this invention lies in constructing a unified modeling and solution framework for the collaborative optimization of unit production and marketing in the electricity spot market. By unifying the modeling of multiple objectives such as generation cost, unit operation stability, market revenue, deviation assessment, and price fluctuation risk, it achieves deep coupling between production-side and marketing-side decision-making. On this basis, a multi-stage energy disturbance collaborative optimization mechanism is introduced. Through individual state energy driving, environmental density adjustment, and adaptive disturbance search, a dynamic balance between the global search and local convergence processes is achieved, thereby effectively avoiding the problem of traditional optimization methods easily getting trapped in local optima. At the same time, by embedding strong physical constraints and market rule constraints into the optimization process, it ensures that the optimization results are engineering-executable under the premise of meeting the requirements of unit operation safety and market compliance.

[0093] The main protected aspects of this invention include: a production-marketing integrated modeling method based on multi-objective weighted or equivalent multi-objective fusion; a collaborative decision variable construction method oriented towards price curves and contract allocation ratios; a collaborative optimization solution method integrating energy-driven mechanisms, density adjustment mechanisms, and disturbance escape mechanisms; a constraint handling mechanism that simultaneously embeds unit physical constraints and market rule constraints during the optimization process; and a complete system architecture based on the above methods and its application in thermal power units, electricity spot markets, and related energy optimization scenarios. These technical features are organically combined to form a unified whole, collectively constituting the core protection scope of this invention.

[0094] See Figure 2 This invention also discloses a system for coordinating and optimizing generator output and pricing in the electricity spot market, comprising: The data preparation module is used to acquire the operating data and market data of thermal power units, and to construct an input data matrix containing time series data based on the operating data and market data. The multi-objective modeling module is used to establish a unified output-price collaborative optimization objective function based on the input data matrix. The objective function includes unit operating cost, market revenue, and risk control, and assigns weight coefficients to each objective. The constraint processing module is used to embed the physical constraints of unit operation and market trading rule constraints as strong constraints into the output-price collaborative optimization objective function; The collaborative optimization and decision output module is used to iteratively solve the output-bid collaborative optimization objective function with embedded strong constraints using a multi-stage energy perturbation collaborative mechanism; when the convergence condition is met, it outputs the collaborative optimization result containing the optimal bid curve and the contract allocation ratio.

[0095] This invention discloses a collaborative optimization system for generating unit output and pricing in the electricity spot market. First, its modular architecture achieves clear decoupling and efficient integration of the production and marketing collaborative optimization processes. The system is divided into modules according to a logical chain of data preparation, multi-objective modeling, constraint handling, collaborative optimization, and decision output. Each module has clearly defined functional boundaries and standardized data interfaces. The data preparation module is responsible for collecting and aligning raw data, providing a unified input benchmark for subsequent modeling. The multi-objective modeling module and the constraint handling module construct the mathematical expression of the optimization problem in parallel, allowing for independent addition, deletion, and adjustment of objective functions and constraints without affecting the core logic of other modules. The collaborative optimization and decision output module serves as the solution engine and is loosely coupled with the front-end modeling module. This architectural design facilitates rapid deployment and maintenance of the system within existing information platforms in thermal power plants and supports module-level replacement or upgrades for different market rules or generating unit characteristics, greatly improving the system's engineering adaptability and scalability. Second, the tight coupling design between the constraint handling module and the multi-objective modeling module ensures the physical feasibility and market compliance of the optimization results. The constraint processing module in this system does not simply filter or penalize solutions in the later stages of optimization. Instead, it directly embeds the physical constraints of unit operation (such as output boundaries and ramp rates) and market trading rule constraints (such as bid limits and contract allocation ratios) as strong constraints into the solution space of the output-bid co-optimization objective function. During the iterative solution process, the co-optimization and decision output module always generates candidate solutions only within the feasible region. This mechanism ensures that the optimal bid curve and contract allocation ratio output by the system meet the requirements for safe unit operation and electricity market declaration rules without any post-processing corrections, eliminating the risk of "feasible but unusable" optimization results in traditional systems and significantly improving the engineering executability of the decision results. Third, the multi-stage energy perturbation coordination mechanism built into the co-optimization and decision output module endows the system with powerful global optimization capabilities and local convergence stability. During the solution process, this module dynamically simulates individual energy levels and environmental density, adaptively switching between global exploration and local development, and automatically executing escape operations when search stagnation is detected. Compared to existing systems employing conventional heuristic algorithms (such as genetic algorithms and particle swarm optimization), this system significantly reduces the probability of getting trapped in local optima when facing high-dimensional, nonlinear, and multi-peak output-bid co-optimization problems, thus achieving better overall returns with the same computational resources. Simultaneously, this module integrates convergence determination and decision output into the same process. Upon meeting the convergence conditions, it directly outputs the bid curve and contract allocation ratio, avoiding additional manual judgment and improving decision response speed. Fourth, the system possesses inherent robustness in handling market uncertainty. The multi-objective modeling module explicitly incorporates risk control terms (such as price volatility variance) when constructing the objective function and weights them with the market return term.This means that when the system performs collaborative optimization, it does not simply pursue maximum returns, but simultaneously weighs the volatility risk of returns. Driven by this risk-return joint objective, the collaborative optimization and decision output module can automatically adjust the shape of the price curve and the contract allocation ratio according to different market volatility levels, for example, proactively reducing risk exposure during periods of high volatility. This endogenous risk handling mechanism enables the system's output decision results to maintain relatively robust overall benefits under different market environments, more closely aligning with the decision-making preferences of thermal power enterprises in actual operation, which balance increasing revenue and mitigating risks. Fifth, the seamless integration between the data preparation module and the front-end module provides high-quality data support for the system. The data preparation module aligns historical output sequences, spot price sequences, fuel cost data, and meteorological influencing factors at the time granularity according to the scheduling cycle (e.g., 96 points), constructing a structured input data matrix. This matrix not only provides a unified time benchmark for the multi-objective modeling module but also facilitates batch vectorized calculations by subsequent modules. Compared with existing systems with scattered and inconsistent data formats, this invention, through a standardized data preparation process, reduces optimization deviations caused by data quality issues, improving the stability of system operation and the reproducibility of results. In summary, the unit output and bidding collaborative optimization system proposed in this invention for the electricity spot market achieves superior technical results compared to existing systems in terms of decision quality, constraint satisfaction capability, global optimization performance, risk robustness, and ease of engineering deployment through reasonable module division and functional collaboration. It provides thermal power companies with a reliable and efficient decision support tool for participating in the electricity spot market.

[0096] A third objective of this invention is to provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the method for co-optimizing generator output and pricing for the electricity spot market.

[0097] The method for coordinating and optimizing unit output and bidding for the electricity spot market includes the following steps: Obtain operating data and market data of thermal power units, and construct an input data matrix containing time series data based on the operating data and market data; Based on the input data matrix, a unified output-price collaborative optimization objective function is established. The objective function includes unit operating cost, market revenue, and risk control items, and weight coefficients are assigned to each item. The physical constraints of unit operation and the constraints of market trading rules are used as strong constraints and embedded into the output-price collaborative optimization objective function; A multi-stage energy perturbation coordination mechanism is used to iteratively solve the output-bid coordination optimization objective function with embedded strong constraints; when the convergence condition is met, the output includes the coordination optimization result containing the optimal bid curve and the contract allocation ratio.

[0098] The fourth objective of this invention is to provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method for co-optimizing unit output and pricing for the electricity spot market.

[0099] The method for coordinating and optimizing unit output and bidding for the electricity spot market includes the following steps: Obtain operating data and market data of thermal power units, and construct an input data matrix containing time series data based on the operating data and market data; Based on the input data matrix, a unified output-price collaborative optimization objective function is established. The objective function includes unit operating cost, market revenue, and risk control items, and weight coefficients are assigned to each item. The physical constraints of unit operation and the constraints of market trading rules are used as strong constraints and embedded into the output-price collaborative optimization objective function; A multi-stage energy perturbation coordination mechanism is used to iteratively solve the output-bid coordination optimization objective function with embedded strong constraints; when the convergence condition is met, the output includes the coordination optimization result containing the optimal bid curve and the contract allocation ratio.

[0100] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0101] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1A device that provides the functions specified in one or more boxes.

[0102] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0103] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0104] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., 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 coordinated optimization of generating unit output and bidding price in the electricity spot market, characterized in that, include: Obtain operating data and market data of thermal power units, and construct an input data matrix containing time series data based on the operating data and market data; Based on the input data matrix, a unified output-price collaborative optimization objective function is established. The objective function includes unit operating cost, market revenue, and risk control items, and weight coefficients are assigned to each item. The physical constraints of unit operation and the constraints of market trading rules are used as strong constraints and embedded into the output-price collaborative optimization objective function; A multi-stage energy perturbation coordination mechanism is used to iteratively solve the output-bid coordination optimization objective function with embedded strong constraints; when the convergence condition is met, the output includes the coordination optimization result containing the optimal bid curve and the contract allocation ratio.

2. The method for coordinated optimization of generating unit output and bidding price for the electricity spot market according to claim 1, characterized in that, The multi-stage energy perturbation coordination mechanism includes: Based on individual energy levels and local environment density, an adaptive global exploration or local exploitation strategy is selected to update candidate solutions; When the optimization process is detected to be trapped in a local optimum, a negative curvature escape operation is performed to escape the local region.

3. The method for coordinated optimization of unit output and bidding price for the electricity spot market according to claim 1, characterized in that, The acquisition of operating data and market data of thermal power units, and the construction of an input data matrix containing time series data based on the operating data and market data specifically include: Basic data of thermal power units within the dispatch cycle are collected, including historical output sequences, spot market price sequences, fuel cost data, and meteorological influencing factors. This data is aligned according to time granularity, and the dispatch cycle is divided into [number] periods. Construct the input matrix at discrete time points: in, For the input data matrix; The number of feature dimensions; This represents the length of the time series.

4. The method for coordinated optimization of unit output and bidding price for the electricity spot market according to claim 1, characterized in that, The unified output-price collaborative optimization objective function established based on the input data matrix includes: Define the output-quote collaborative optimization objective function as follows: in, To optimize the objective function value for output-quote collaboration, - The target weight coefficient; The sub-objectives are defined as follows: Fuel cost item: in, This is the unit fuel cost coefficient for electricity generation. For a moment The unit output; Start-up and shutdown costs: in, This refers to the start-up and shutdown cost coefficient. This is a unit status variable, where 0 indicates shutdown and 1 indicates operation; Climbing stability: in, This is the climbing penalty coefficient; Market returns: in, For a moment The quote; Deviation assessment: in, For planned or settled electricity consumption; This is the deviation penalty coefficient; risk: in, Let V be the price volatility variance.

5. The method for coordinated optimization of generating unit output and bidding price for the electricity spot market according to claim 1, characterized in that, The inclusion of physical constraints on unit operation and market trading rules as strong constraints in the output-price collaborative optimization objective function includes: The physical constraints and market trading rules governing the operation of the generating units include: Output boundary constraints: Climbing constraints: Quotation constraints: Contract allocation constraint: Assume the contract allocation ratio is... ,satisfy in, These are the unit's minimum and maximum outputs; It is the maximum climbing rate; These are the upper and lower limits of market prices.

6. The method for coordinated optimization of unit output and bidding price for the electricity spot market according to claim 1, characterized in that, The iterative solution of the output-bid co-optimization objective function with embedded strong constraints using a multi-stage energy perturbation cooperative mechanism includes: Candidate solution initialization: Let the population size be Each candidate solution is represented as: in, For the price quote sequence vector, Assign a vector to the contract; Energy-driven search mechanism: Define individual energy: in, For the first The individual in the first Energy of the next iteration; This is the distance consumption coefficient; Update the distance to the location; Return enhancement factor; The improvement amount of the objective function; Environmental density regulation mechanism: Define local density as The adjustment factor is: Used to control the search step size; Adaptive position update: when Perform a global update at the specified time: Otherwise, perform a partial update: in, This is the current optimal solution; For random neighbor solutions; This is the iteration step size factor; Negative curvature escape mechanism: When a negative local curvature is detected: in, The direction is random; The escape amplitude is used to avoid getting trapped in local optima. Random exploration and smoothing mechanism: Introduce random perturbations: And perform exponential smoothing: in, Step size; The disturbance intensity; It is random noise; This is the smoothing coefficient.

7. The method for coordinated optimization of unit output and bidding price for the electricity spot market according to claim 1, characterized in that, When the convergence condition is met, the output includes the co-optimization result containing the optimal bid curve and the contract allocation ratio, including: Once the iteration converges, the optimal decision result is output: in, This represents the optimal pricing curve. To achieve the optimal contract allocation ratio; The optimal decision result is directly used for market application and production scheduling; After actual operation, obtain the actual clearing results, deviation assessment results, and profit realization status, and update the target weights. Risk parameters and optimized initial population are used to form a continuous optimization iteration.

8. A system for coordinated optimization of generating unit output and bidding price for the electricity spot market, characterized in that, include: The data preparation module is used to acquire the operating data and market data of thermal power units, and to construct an input data matrix containing time series data based on the operating data and market data. The multi-objective modeling module is used to establish a unified output-price collaborative optimization objective function based on the input data matrix. The objective function includes unit operating cost, market revenue, and risk control, and assigns weight coefficients to each objective. The constraint processing module is used to embed the physical constraints of unit operation and market trading rule constraints as strong constraints into the output-price collaborative optimization objective function; The collaborative optimization and decision output module is used to iteratively solve the output-bid collaborative optimization objective function with embedded strong constraints using a multi-stage energy perturbation collaborative mechanism; when the convergence condition is met, it outputs the collaborative optimization result containing the optimal bid curve and the contract allocation ratio.

9. An electronic device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the unit output and bidding co-optimization method for the electricity spot market as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the unit output and bidding co-optimization method for the electricity spot market as described in any one of claims 1-7.

Images