A peak shaving auxiliary service cooperation game cost allocation method and system considering environmental externality

Abstract

The present application relates to the technical field of power system operation optimization, and particularly relates to a peak shaving auxiliary service cooperation game cost allocation method and system considering environmental externality, which comprises obtaining system time series data and unit operation emission data, determining the net load fluctuation of a peak shaving demand subject according to the system time series data, and generating an environmental cost item representing the environmental externality of deep peak shaving of a thermal power unit according to the unit operation emission data.The present application determines the net load fluctuation of a peak shaving demand subject and the environmental cost item representing the environmental externality of deep peak shaving of a thermal power unit by obtaining system time series data and unit operation emission data, and introduces the environmental cost item into the dispatch optimization target of joint clearing, so that the joint clearing model can simultaneously consider the system operation cost and the environmental cost generated by deep peak shaving when optimizing the dispatch strategy.

Description

Technical Field

[0001] This invention relates to the field of power system operation optimization technology, and in particular to a method and system for cost sharing in cooperative game theory for peak-shaving ancillary services that takes into account environmental externalities. Background Technology

[0002] As the proportion of renewable energy sources such as wind and solar power integrated into the power system continues to increase, the fluctuation range of the system's net load is widening. Flexible resources such as thermal power units and pumped storage need to participate in peak-shaving ancillary services more frequently to maintain the power system's supply-demand balance and safe, stable operation. In the existing settlement process for peak-shaving ancillary services, cost allocation is usually determined primarily based on factors such as peak-shaving power volume, load scale, or market quotations, with less consideration given to the actual impact of time-series power fluctuations of different peak-shaving demand entities on the system's peak-shaving pressure. This results in a lack of accurate correlation between peak-shaving costs and the entities that trigger peak-shaving demand.

[0003] Meanwhile, thermal power units typically operate at low loads during deep peak shaving, resulting in decreased unit efficiency and potentially additional coal consumption, carbon emissions, and pollution control costs associated with oil injection for peak shaving. Existing joint clearing or ancillary service cost-sharing methods often separate peak shaving compensation costs from environmental externalities, making it difficult to simultaneously reflect system operating costs and the environmental costs of deep peak shaving during dispatch optimization. Consequently, the cost-sharing ancillary services cost-sharing results fail to fully reflect the actual marginal impact of each peak shaving demand entity on system peak shaving costs and environmental costs.

[0004] Therefore, how to introduce the environmental externality costs generated by deep peak shaving of thermal power units during the joint clearing process, and how to strip away the dynamic incremental costs based on net load fluctuations, and further form a fair and traceable cost allocation result for peak shaving ancillary services based on the marginal contributions of each peak shaving demand subject, has become a technical problem that needs to be solved. Summary of the Invention

[0005] In this section, as well as in the abstract and title of this application, some simplifications or omissions may be made to avoid obscuring the purpose of this section, the abstract, and the title of this application, and such simplifications or omissions shall not be used to limit the scope of the invention.

[0006] To address the shortcomings of existing technologies, one objective of this invention is to provide a method for cost sharing in a cooperative game theory approach for peak-shaving ancillary services that takes into account environmental externalities.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: a peak-shaving auxiliary service cooperative game cost-sharing method considering environmental externalities, comprising: The system acquires time-series data and unit operation emission data. Based on the system time-series data, it determines the net load fluctuation of the peak-shaving demand entities and generates an environmental cost item characterizing the environmental externalities of deep peak-shaving by thermal power units based on the unit operation emission data. It defines scheduling constraints, incorporates the environmental cost item into the joint clearing scheduling optimization objective, and uses the scheduling optimization objective as the objective function and the scheduling constraints as the constraint conditions. A mixed-integer optimization model considering environmental costs is constructed. The scheduling strategy is optimized through the joint clearing model, and the dynamic incremental cost caused by the net load fluctuation is extracted as a characteristic function value. Based on the characteristic function value, the marginal contribution of each peak-shaving demand entity is determined, and a peak-shaving ancillary service cost allocation result is generated based on the marginal contribution.

[0008] As a preferred embodiment of the peak-shaving auxiliary service cooperative game cost-sharing method considering environmental externalities described in this invention, the system time-series data includes the time-series power curves of the peak-shaving demand subjects, the time-series power curves include the actual output curves of wind power, photovoltaic power, and actual load curves of users; the unit operation emission data includes environmental price parameters and unit operating condition parameters, the environmental price parameters include carbon price, sulfur dioxide environmental treatment unit price, and nitrogen oxide environmental treatment unit price, the unit operating condition parameters include unit output, unit capacity level, deep peak-shaving level, minimum output for deep peak-shaving without oil injection, and coal consumption increment.

[0009] As a preferred embodiment of the peak-shaving auxiliary service cooperative game cost-sharing method considering environmental externalities described in this invention, the following steps are taken: When determining the net load fluctuation, the actual user load curve, the actual wind power output curve, and the actual photovoltaic output curve are unified to the same time scale; a load deviation component is obtained based on the deviation of the actual user load curve from the corresponding load average baseline; a new energy output deviation component is obtained based on the deviation of the actual wind power output curve and the actual photovoltaic output curve from the corresponding new energy output average baseline; the load deviation component is used as a positive component that increases the system net load, and the new energy output deviation component is used as a negative component that reduces the system net load; the load deviation component and the new energy output deviation component are synthesized to obtain the net load fluctuation.

[0010] As a preferred embodiment of the peak-shaving auxiliary service cooperative game cost-sharing method considering environmental externalities described in this invention, the method comprises: when generating the environmental cost item, determining the peak-shaving operation range of the thermal power unit based on the unit operating parameters, and obtaining the emission increment parameters corresponding to the peak-shaving operation range; determining the emission conversion price based on the environmental price parameters; using the paid peak-shaving benchmark value as the calculation benchmark, calculating the additional carbon emission cost caused by the increase in coal combustion and the pollution gas treatment cost caused by the increase in oil production based on the emission increment parameters and the emission conversion price; and combining the additional carbon emission cost and the pollution gas treatment cost into the environmental cost item.

[0011] As a preferred embodiment of the peak-shaving auxiliary service cooperative game cost-sharing method considering environmental externalities described in this invention, the scheduling constraints include system balance constraints, system reserve constraints, thermal power unit operation constraints, renewable energy output constraints, and pumped storage operation constraints; the system balance constraints are used to constrain the balance relationship between thermal power output, renewable energy output, pumped storage power, and system load demand; the system reserve constraints are used to constrain the system's positive and negative spinning reserve capacity; the thermal power unit operation constraints are used to constrain the unit's operating status, maximum technical output, minimum technical output, upward ramp rate, downward ramp rate, minimum continuous operating time, and minimum continuous downtime; the renewable energy output constraints are used to constrain wind power output and photovoltaic output to not exceed the corresponding predicted output; the pumped storage operation constraints are used to constrain the mutual exclusion of the pumping and power generation states of the pumped storage power station in the same period, and to constrain the water volume status of the upper and lower reservoirs and the energy conservation relationship.

[0012] As a preferred embodiment of the peak-shaving auxiliary service cooperative game cost-sharing method considering environmental externalities described in this invention, the following steps are taken: when constructing the joint clearing model, the scheduling optimization objective is set as the objective of minimizing the total system operating cost; the scheduling optimization objective is synthesized from the basic electricity supply cost, thermal power unit start-up and shutdown cost, deep peak-shaving compensation cost, pumped storage operation cost, renewable energy reduction cost, and the environmental cost item; the scheduling optimization objective is used as the objective function of the joint clearing model, and the scheduling constraints are used as the constraints of the joint clearing model; the joint clearing model is optimized and solved to obtain a scheduling strategy that satisfies the scheduling constraints.

[0013] As a preferred embodiment of the peak-shaving auxiliary service cooperative game cost-sharing method considering environmental externalities described in this invention, the method comprises: when determining the characteristic function value, replacing the time-series power curves of each peak-shaving demand subject with the corresponding mean line to form a baseline substitution scenario with no peak-shaving demand; calling the joint clearing model to optimize and clear the baseline substitution scenario to obtain the baseline clearing cost; for any combination of peak-shaving demand subjects, generating a combined net load fluctuation based on the deviation of the time-series power curves of each peak-shaving demand subject within the combination from the corresponding mean line; superimposing the combined net load fluctuation onto the baseline substitution scenario to form a coalition peak-shaving scenario; calling the joint clearing model to optimize and clear the coalition peak-shaving scenario to obtain the combined clearing cost; and determining the characteristic function value corresponding to the combination of peak-shaving demand subjects based on the difference between the combined clearing cost and the baseline clearing cost.

[0014] As a preferred embodiment of the peak-shaving auxiliary service cooperative game cost-sharing method considering environmental externalities described in this invention, the method comprises: when determining the marginal contribution, using the Shapley value rule to calculate the weighted average marginal contribution of the target entity in different combinations of peak-shaving demand entities; wherein, for any target entity, the difference in characteristic function values ​​between the combination of peak-shaving demand entities including the target entity and the combination of peak-shaving demand entities excluding the target entity is calculated to obtain the marginal cost increment of the target entity under the corresponding combination; the marginal cost increment of the target entity under different combinations of peak-shaving demand entities is weighted to obtain the marginal contribution of the target entity.

[0015] As a preferred embodiment of the peak-shaving auxiliary service cooperative game cost allocation method considering environmental externalities described in this invention, the method comprises: when generating the peak-shaving auxiliary service cost allocation result, determining the peak-shaving auxiliary service cost to be allocated based on the characteristic function value corresponding to the combination of peak-shaving demand subjects including all the peak-shaving demand subjects; determining the corresponding allocation ratio based on the marginal contribution of each peak-shaving demand subject; allocating the peak-shaving auxiliary service cost to be allocated according to the allocation ratio to obtain the allocation cost corresponding to each peak-shaving demand subject; generating the peak-shaving auxiliary service cost allocation result based on each allocation cost; wherein, when the marginal contribution of a peak-shaving demand subject indicates that it has a reducing effect on the system's peak-shaving cost, the allocation cost of that peak-shaving demand subject is reduced.

[0016] To address the shortcomings of existing technologies, another objective of this invention is to provide a peak-shaving auxiliary service cooperative game cost-sharing system that takes into account environmental externalities.

[0017] This invention adopts the following technical solution: a peak-shaving auxiliary service cooperative game cost-sharing system that takes into account environmental externalities, comprising: A data processing unit is used to acquire system time-series data and unit operation emission data, and generate net load fluctuation and environmental cost items based on the system time-series data and the unit operation emission data, respectively. A joint clearing unit is connected to the data processing unit, used to receive the net load fluctuation and the environmental cost items, and construct a joint clearing model based on scheduling constraints and scheduling optimization objectives that incorporate the environmental cost items, so as to determine the characteristic function value through the joint clearing model. A cost allocation unit is connected to the joint clearing unit, used to receive the characteristic function value, determine the marginal contribution of each peak-shaving demand subject based on the characteristic function value, and generate the peak-shaving ancillary service cost allocation result based on the marginal contribution.

[0018] The beneficial effects of the peak-shaving ancillary service cooperative game cost-sharing method considering environmental externalities of the present invention are as follows: The present invention obtains system time-series data and unit operation emission data to determine the net load fluctuation of peak-shaving demand subjects and the environmental cost item characterizing the environmental externalities of deep peak-shaving by thermal power units, respectively. This environmental cost item is then introduced into the joint clearing scheduling optimization objective, enabling the joint clearing model to simultaneously consider system operating costs and the environmental costs generated by deep peak-shaving when optimizing scheduling strategies. Furthermore, the present invention uses the joint clearing model to remove the dynamic incremental cost caused by net load fluctuations, treating it as a characteristic function value. Based on this characteristic function value, the marginal contribution of each peak-shaving demand subject is determined, thereby corresponding the peak-shaving ancillary service cost to the actual impact of each subject on the system's peak-shaving pressure. Therefore, the present invention can improve the fairness and interpretability of the peak-shaving ancillary service cost-sharing results, ensuring that the subject causing the increase in system peak-shaving costs bears the corresponding cost, and that the subject that reduces system peak-shaving costs receives corresponding recognition. This is conducive to forming a cost-sharing mechanism that balances economic scheduling and environmental responsibility transmission. Attached Figure Description

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

[0020] Figure 1 This is a schematic diagram of the overall process of the peak-shaving auxiliary service cooperative game cost-sharing method that takes into account environmental externalities in an embodiment of the present invention.

[0021] Figure 2 This is a schematic diagram showing the system timing data and the timing power curve of the main peak-shaving demand in an embodiment of the present invention.

[0022] Figure 3 This is a schematic diagram illustrating the composition of unit operation emission data in an embodiment of the present invention.

[0023] Figure 4 This is a schematic diagram illustrating the composition of environmental price parameters in an embodiment of the present invention.

[0024] Figure 5 This is a schematic diagram illustrating the composition of the unit operating parameters in an embodiment of the present invention.

[0025] Figure 6 This is a schematic diagram illustrating the generation process of net load fluctuations in an embodiment of the present invention.

[0026] Figure 7 This is a schematic diagram illustrating the generation process of the environmental cost item in an embodiment of the present invention.

[0027] Figure 8 This is a schematic diagram illustrating the composition of scheduling constraints in an embodiment of the present invention.

[0028] Figure 9 This is a schematic diagram illustrating the construction and clearing process of the joint clearing model in an embodiment of the present invention.

[0029] Figure 10 This is a schematic diagram illustrating the process of determining the characteristic function value in an embodiment of the present invention.

[0030] Figure 11 This is a schematic diagram illustrating the process of determining marginal contribution in an embodiment of the present invention.

[0031] Figure 12 This is a schematic diagram illustrating the process of generating the cost allocation result for peak shaving ancillary services in an embodiment of the present invention.

[0032] Figure 13 This is a schematic diagram of the cost-sharing system for peak-shaving auxiliary services that takes into account environmental externalities in an embodiment of the present invention.

[0033] Figure 14 This is a schematic diagram of the deep peak shaving level in an embodiment of the present invention.

[0034] Figure 15 This invention provides a per-unit curve of peak demand and a load forecast curve.

[0035] Figure 16 This is a deep peak shaving damper diagram for a unit in an embodiment of the present invention.

[0036] Figure 17 This is another deep peak shaving chute diagram in an embodiment of the present invention.

[0037] In the diagram: 10. System time-series data; 101. Time-series power curve; 11. Unit operation emission data; 111. Environmental price parameters; 112. Unit operating condition parameters; 12. Net load fluctuation; 121. Load deviation component; 122. Renewable energy output deviation component; 123. Combined net load fluctuation; 13. Environmental cost item; 131. Peak-shaving operation range; 132. Emission increment parameters; 133. Emission conversion price; 134. Paid peak-shaving benchmark value; 135. Additional carbon emission cost; 136. Pollutant gas treatment cost; 14. Dispatch constraints; 141. System balance constraints; 142. System reserve constraints; 143. Thermal power unit operation constraints; 144. Renewable energy output constraints; 145. Pumped storage operation constraints; 15. Dispatch optimization objectives; 151. Base electricity. 152. Supply cost; 153. Start-up and shutdown cost of thermal power units; 154. Deep peak shaving compensation cost; 155. Pumped storage operation cost; 16. Renewable energy reduction cost; 17. Joint clearing model; 18. Dispatch strategy; 19. Clearing cost; 10. Characteristic function value; 11. Baseline clearing cost; 11. Combined clearing cost; 12. Marginal contribution; 13. Marginal cost increment; 14. Peak shaving ancillary service cost allocation result; 156. Peak shaving ancillary service cost to be allocated; 177. Allocation ratio; 18. Allocation cost; 198. Baseline substitution scenario; 20. Alliance peak shaving scenario; 21. Peak shaving demand subject; 222. Peak shaving demand subject combination; 223. Target subject; 100. Data processing unit; 200. Joint clearing unit; 300. Cost allocation unit. Detailed Implementation

[0038] To enable those skilled in the art to better understand the present invention, the present invention will be further described in detail below with reference to specific embodiments and accompanying drawings.

[0039] The terminology used in this invention is that which is currently widely used in the art in consideration of the function of the invention; however, these terms may vary according to the intent of those skilled in the art, precedent, or new technology in the art. Furthermore, specific terms may be chosen by the applicant, and in such cases, their detailed meanings will be described in the detailed description of the invention. Therefore, the terms used in this specification should not be construed as simple names, but rather based on their meanings and the overall description of the invention.

[0040] Example 1

[0041] Reference Figure 1This embodiment provides a cooperative game-theoretic cost-sharing method for peak-shaving ancillary services that takes into account environmental externalities. This method can be executed by a power dispatch center server, a power market clearing server, an ancillary service settlement platform, or a computing device equipped with an optimization solver. The inputs to this method include system time-series data 10 and unit operation emission data 11, and the output is the peak-shaving ancillary service cost-sharing result 19.

[0042] System time-series data 10 describes the actual output or load demand changes of peak-shaving demand entity 22 within the scheduling cycle. Unit operation emission data 11 describes the operating status, coal consumption changes, oil injection status, and emission conversion prices of thermal power units under different peak-shaving conditions. Peak-shaving demand entity 22 can be divided according to market entities, user types, power source types, or aggregation units. For example, residential load, manufacturing load, service industry load, wind farms, and photovoltaic power plants can be classified as different peak-shaving demand entities 22. The purpose of this division is to ensure that subsequent marginal contributions 18 correspond to entities with settlement significance, rather than merely remaining at the system total level.

[0043] In the first phase, system time-series data 10 and unit operation emission data 11 are acquired. System time-series data 10 can come from the dispatch system database, energy management system, market trading system, or new energy power acquisition system. Unit operation emission data 11 can come from the thermal power unit operation monitoring system, environmental monitoring platform, carbon trading system, unit economic test data, or unit peak-shaving quotation data. To ensure that the data can be used for joint clearing and cooperative game-based cost allocation, data from different sources and sampling intervals need to be unified to the same dispatch period during data processing, for example, to a time scale of 15 minutes, 30 minutes, or 1 hour.

[0044] In the second stage, the net load fluctuation 12 of the peak-shaving demand entity 22 is determined based on the system time-series data 10. Ordinary cost allocation methods typically only consider the amount of electricity generated by the entity within the settlement period, but the amount of electricity generated does not necessarily equate to peak-shaving responsibility. For example, a load entity with a large total electricity generation but a stable load curve may have a lower impact on the system's peak-shaving pressure than an entity with a smaller total electricity generation but a more pronounced peak-to-valley difference. This embodiment uses the net load fluctuation 12 as the basis for subsequent dynamic incremental cost stripping, thus linking cost allocation to time-series fluctuations.

[0045] Specifically, during data processing, the deviation of the load-side curve from the load average baseline and the deviation of the output curves of wind power, photovoltaic power, and other new energy sources from the new energy output average baseline can be calculated separately. Load deviation manifests as positive peak-shaving pressure when increasing the system's net load; new energy output deviation manifests as a reverse component when reducing the system's net load. By synthesizing the positive and reverse components, the net load fluctuations 12 caused by peak-shaving demand entities 22 in each time period can be obtained. Compared to directly using the original power curves, this processing method can distinguish between "the entity's own scale" and "the dynamic fluctuations caused by the entity," making it more suitable as a basis for tracking ancillary service cost responsibilities.

[0046] In the third stage, an environmental cost item 13 characterizing the environmental externalities of deep peak shaving for thermal power units is generated based on the unit operation emission data 11. When thermal power units operate within the basic peak shaving range, coal consumption is relatively stable, typically without significant additional environmental costs. However, during deep peak shaving, unit load rates decrease, efficiency declines, and coal consumption per unit of electricity increases. In some extremely deep peak shaving conditions, oil injection may be necessary for stable combustion, resulting in additional carbon emission costs and treatment costs for pollutants such as sulfur dioxide and nitrogen oxides. This embodiment treats this environmental cost as environmental cost item 13, rather than processing it as an external additional charge after the clearing process is complete.

[0047] The technical significance of this setup is that after environmental cost item 13 is introduced into the joint clearing scheduling optimization objective 15, the joint clearing model 16 will simultaneously compare economic and environmental costs when selecting scheduling strategies such as deep peak shaving of thermal power units, pumped storage regulation, and renewable energy reduction. For example, if the environmental cost caused by extremely deep peak shaving is significantly higher than the cost of moderate wind curtailment or pumped storage in a certain period, the model may tend to choose the scheduling strategy with lower overall cost, thus making scheduling strategy 161 closer to minimizing the actual social cost of the system.

[0048] In the fourth stage, scheduling constraints 14 are defined, and environmental cost item 13 is introduced into the joint clearing scheduling optimization objective 15. Using scheduling optimization objective 15 as the objective function and scheduling constraints 14 as the constraint conditions, a joint clearing model 16 incorporating environmental costs is constructed using a mixed-integer optimization modeling method. The joint clearing model 16 can be a mixed-integer linear programming model, or it can be configured as a mixed-integer nonlinear programming model or an optimized model after linearization, depending on the actual system requirements. The mixed-integer optimization modeling method is advantageous because it accommodates discrete variables such as unit start-up and shutdown states, pumped storage pumping / generation mutual exclusion states, and deep peak-shaving selection, while simultaneously handling continuous variables such as unit output, peak-shaving volume, and reservoir water volume.

[0049] The scheduling optimization objective 15 can be set as minimizing the total system operating cost. This objective is not simply an electricity cost objective, but rather a synthesis of the basic electricity supply cost, thermal power unit start-up and shutdown cost, deep peak shaving compensation cost, pumped storage operation cost, renewable energy reduction cost, and environmental cost item 13. Therefore, the joint clearing model 16 considers both the economics of "who generates electricity and who shaves peaks" and the environmental externalities caused by deep peak shaving, thus avoiding the problem that the existing energy market and ancillary service market separation clearing cannot accurately reflect the social costs of deep peak shaving.

[0050] In the fifth stage, the scheduling strategy is optimized through the joint clearing model 16, and the dynamic incremental cost caused by net load fluctuations 12 is extracted as the characteristic function value 17. Specifically, a baseline replacement scenario 20 without peak-shaving demand can be constructed, in which the time-series power curves of each peak-shaving demand subject 22 are replaced with the corresponding mean line, so that the system only needs to meet the relatively stable power balance demand. Then, for different combinations of peak-shaving demand subjects 221, the fluctuation deviations of each subject in the combination relative to the mean line are superimposed on the baseline replacement scenario 20 to form the alliance peak-shaving scenario 21. By optimizing and clearing the baseline replacement scenario 20 and the alliance peak-shaving scenario 21 through the joint clearing model 16, the incremental cost caused by the fluctuation of this combination can be obtained.

[0051] The aforementioned incremental cost, as characteristic function value 17, is essentially a value function used in cooperative game theory to describe the impact of a certain combination of peak-shaving demand subjects 221 on the system's peak-shaving cost pool. Unlike directly allocating the total system cost proportionally, this embodiment, through a comparison of "benchmark substitution scenario 20 - alliance peak-shaving scenario 21," separates fixed costs or basic operating costs unrelated to the fluctuations of peak-shaving demand subjects, retaining only the dynamic incremental costs triggered by net load fluctuations 12. This prevents stable subjects from bearing costs unrelated to their fluctuations and also identifies situations where certain subjects alleviate the system's peak-shaving pressure.

[0052] In the sixth stage, the marginal contribution 18 of each peak-shaving demand entity 22 is determined based on the characteristic function value 17, and the peak-shaving auxiliary service cost allocation result 19 is generated based on the marginal contribution 18. In the context of cooperative game theory, the marginal contribution 18 of a certain entity can be understood as the weighted average of the changes in the characteristic function value 17 after adding or removing that entity in different peak-shaving demand entity combinations 221 scenarios. When the marginal contribution 18 is positive, it indicates that the temporal fluctuation of that entity increases the overall peak-shaving cost of the system; when the marginal contribution 18 is negative, it indicates that the temporal fluctuation of that entity has the effect of peak shaving and valley filling or smoothing net load fluctuations in some periods, and can be used as a mitigating factor in the allocation process.

[0053] This embodiment offers three improvements over conventional cost-sharing methods based on electricity consumption or capacity. First, the cost pool is derived from a joint clearing model 16 that takes into account environmental costs, reflecting the true social costs of deep peak shaving. Second, cost stripping relies on net load fluctuations 12 and benchmark substitution scenarios 20, distinguishing dynamic incremental costs from basic operating costs. Third, the allocation is based on marginal contribution 18, reflecting the average impact of peak-shaving demand entities 22 across all possible alliance combinations.

[0054] Example 2

[0055] Reference Figures 2 to 5 The system time-series data 10 includes the time-series power curve 101 of the peak-shaving demand subject 22. The time-series power curve 101 may include the actual output curve of wind power, the actual output curve of photovoltaic power, and the actual load curve of users. All of the above curves have time attributes and can reflect the power change direction, fluctuation amplitude, and peak and valley positions of the subject within the dispatch cycle.

[0056] In one specific implementation, the actual wind power output curves and actual photovoltaic power output curves can be collected by the new energy power station monitoring and control system, power prediction system, or dispatch automation system; the actual user load curves can be obtained by the electricity consumption information collection system, dispatch system database, or load aggregation platform. For data from different sources, outlier cleaning, missing value completion, and time alignment can be performed first. For example, data points that significantly exceed the installed capacity or contracted capacity can be removed or corrected by combining data from adjacent time periods; for short-term missing data, linear interpolation, compensation using similar daily curves, or prediction curve correction methods can be used to complete the data.

[0057] The unit operation emission data 11 includes environmental price parameter 111 and unit operating condition parameter 112. Environmental price parameter 111 includes the carbon price, the unit price for sulfur dioxide environmental treatment, and the unit price for nitrogen oxide environmental treatment, used to convert the emission increment into monetary cost. Unit operating condition parameter 112 includes unit output, unit capacity level, deep peak shaving level, minimum output for deep peak shaving without oil injection, and coal consumption increment. These parameters can be collected in real time through the EMS system, or provided by unit performance test reports, peak shaving quotation documents, or environmental monitoring systems.

[0058] Unit capacity class can be used to differentiate the peak-shaving capacity and coal consumption increment characteristics of different types of units. For example, the briefing uses 600MW and 300MW units as typical units, and divides the peak-shaving of thermal power units into multiple levels according to the ratio of rated capacity PN. The coal consumption increment, oil injection conditions, and pollutant emission levels of units of different capacity classes may differ under the same peak-shaving depth level. Therefore, incorporating the unit capacity class into the unit operating condition parameter 112 can improve the calculation accuracy of the environmental cost item 13.

[0059] In this embodiment, the system time-series data 10 and the unit operation emission data 11 are not isolated inputs, but jointly serve the generation of two subsequent key objects: net load fluctuation 12 and environmental cost item 13. The former is used to identify "which entity triggered the peak-shaving demand," and the latter is used to quantify "what environmental cost the system needs to pay to meet this peak-shaving demand." Through this data structuring method, peak-shaving responsibility and environmental responsibility can be passed on in the same technological chain.

[0060] Example 3

[0061] Reference Figure 6 This embodiment further illustrates the method for determining net load fluctuation 12 based on Embodiment 2. To avoid inconsistencies in curve sampling periods, metering calibers, and time stamps among different entities, the actual user load curve, actual wind power output curve, and actual photovoltaic output curve are unified to the same time scale before determining net load fluctuation 12. After unifying the time scale, each scheduling period corresponds to a load value, one or more new energy output values, and a corresponding entity identifier.

[0062] For the actual load curve of a user, its average load within the scheduling cycle can be calculated first to form a corresponding load average baseline. When the actual load of a user is higher than its load average baseline during a certain period, its deviation can be regarded as a load deviation component that increases the net load of the system. When the actual load of a user is lower than its load average baseline during a certain period, its deviation can reduce the increase in net load of that subject during that period. Through this processing method, the total power volume of the load subject can be distinguished from the fluctuation characteristics of the load curve.

[0063] For the actual output curves of wind power and photovoltaic power, the average baseline of renewable energy output during the dispatch cycle can be calculated separately. When the actual output of wind power or photovoltaic power is higher than the average baseline, its output helps to reduce the net load of the system; when its actual output is lower than the average baseline, it may increase the regulation demand of conventional units. Based on the above deviation, the renewable energy output deviation component 122 can be obtained. This renewable energy output deviation component 122 serves as the reverse component to reduce the net load of the system when synthesizing the net load fluctuation 12.

[0064] In one specific calculation method, the net load fluctuation 12 can be obtained for each time period t by subtracting the new energy output deviation component 122 from the load deviation component 121. To accommodate different types of entities, the deviations of each entity can be normalized or standardized first, and then synthesized according to the entity type and direction. It should be noted that standardization is not a necessary step; when the capacity caliber and unit of measurement are standardized, MW or MWh can be directly used as the calculation unit for the net load fluctuation 12.

[0065] In this embodiment, net load fluctuation 12 serves as a crucial intermediate quantity for both joint clearing and cost allocation. For joint clearing, net load fluctuation 12 determines how much deep peak-shaving resources, pumped storage resources, or energy curtailment measures the system needs to utilize. For cost allocation, net load fluctuation 12 determines the impact of different peak-shaving demand entities 22 on the characteristic function value 17 in different alliance peak-shaving scenarios 21. Thus, net load fluctuation 12 acts as a bridge connecting physical scheduling and market allocation.

[0066] Example 4

[0067] Reference Figure 7 This embodiment further illustrates the generation process of environmental cost item 13. When generating environmental cost item 13, the peak-shaving operating range 131 of the thermal power unit is first determined based on the unit operating condition parameters 112, and the emission increment parameters 132 corresponding to the peak-shaving operating range 131 are obtained. The peak-shaving operating range 131 can be divided according to the unit load rate. For example, 50%PN can be used as the paid peak-shaving benchmark value 134, with operating ranges higher than or equal to 50%PN designated as the basic peak-shaving stage, and operating ranges lower than 50%PN designated as the deep peak-shaving stage.

[0068] In one specific implementation, the peak-shaving operation range 131 can be set with four levels: the first level is 50%-100%PN, corresponding to the basic peak-shaving range; the second level is 45%-50%PN; the third level is 40%-45%PN; and the fourth level is 35%-40%PN. The second and third levels can be peak-shaving ranges without oil injection, mainly reflecting the additional carbon emission costs 135 caused by the increase in coal consumption; the fourth level can be peak-shaving range with oil injection, considering not only the increase in coal consumption but also the pollution gas treatment costs 136 caused by the increase in oil injection. This level division can be adjusted according to the actual power grid peak-shaving rules, unit test results, or market pricing rules.

[0069] The emission increment parameter 132 may include coal consumption increment, carbon emission factor of coal-fired power generation, sulfur dioxide emission increment under oil injection conditions, and nitrogen oxide emission increment. In one specific implementation, the essence of the environmental cost of deep peak shaving is the additional carbon emission cost and pollutant gas treatment cost generated during the deep peak shaving process. Emission calculation can be divided into two categories: incremental coal-fired emissions and pollutant gas treatment costs. Among them, oil injection mainly corresponds to high-level deep peak shaving conditions.

[0070] The emission conversion price 133 is determined based on environmental price parameter 111. The emission conversion price 133 may include the carbon price, the unit price for sulfur dioxide environmental treatment, and the unit price for nitrogen oxide environmental treatment. The additional carbon emission cost 135 can be calculated based on the increase in coal consumption, peak-shaving electricity, carbon emission factor, and carbon price; the pollution gas treatment cost 136 can be calculated based on the increase in sulfur dioxide and nitrogen oxide emissions caused by oil injection for peak shaving, and their treatment unit prices. To facilitate matching with peak-shaving volume, emissions can be uniformly converted into the cost corresponding to a unit of peak-shaving volume.

[0071] For example, the following textual calculation relationships can be used: Carbon emissions per unit of peak-shaving capacity = Increased coal consumption × Carbon emission factor of coal-fired power generation ÷ 1000; Additional carbon emission cost 135 = Carbon emissions per unit of peak-shaving capacity × Peak-shaving capacity × Carbon price; Pollutant gas treatment cost 136 = Increased emissions from oil-fired peak-shaving × Corresponding pollutant treatment unit price. The parameters in the above calculation formulas can be selected according to the unit capacity level, peak-shaving level, and output level.

[0072] After combining the additional carbon emission cost 135 and the pollutant treatment cost 136 into an environmental cost item 13, environmental cost item 13 can be input into the joint clearing model 16 as a component of the scheduling optimization objective 15. Conventional methods often only calculate environmental costs after clearing, making it difficult to influence scheduling strategies. This embodiment internalizes environmental costs into the clearing objective, allowing the model to compare the environmental costs of different peak-shaving strategies during the optimization phase, thus better reflecting the integration of low-carbon scheduling and ancillary service cost sharing.

[0073] The key parameters for environmental cost item 13 are shown in Table 1.

[0074] Table 1: Key Parameters for Environmental Cost Item 13

[0075] Emissions need to be calculated separately for "carbon emissions from increased coal combustion" and "polluting gas emissions from increased oil production," and ultimately unified as follows: To match the peak shaving unit.

[0076] Incremental carbon emissions from coal combustion: Carbon emissions per unit of peak-shaving capacity = incremental coal consumption × carbon emission factor ÷ 1000.

[0077] , in, For the unit In peak shaving mode Carbon emissions from coal combustion per unit of peak-shaving volume; For the unit The increase in coal consumption; The carbon emission factor is the carbon emission factor per unit of standard coal.

[0078] Increased oil injection emissions: The pollutant emissions per unit of peak shaving capacity are the unit's Output during oil injection peak shaving phase and minimum output during deep peak shaving without oil injection The difference in unit output brought about and Emissions: The lower the unit output, the more pollutants are emitted. See the diagram for specific peak-shaving settings. Figure 14 .

[0079] , Deep peak shaving environmental cost calculation: The environmental cost for a single unit, single time period, and single gear is: .

[0080] Example 5

[0081] Reference Figure 8 This embodiment illustrates the specific composition of scheduling constraint 14. Scheduling constraint 14 includes system balance constraint 141, system reserve constraint 142, thermal power unit operation constraint 143, renewable energy output constraint 144, and pumped storage operation constraint 145. These constraints collectively define the feasible region of the joint clearing model 16, ensuring that the optimization results satisfy both the physical operation requirements of the power system and the clearing logic of peak-shaving ancillary services.

[0082] System balance constraint 141 is used to constrain the balance between thermal power output, renewable energy output, pumped storage power, and system load demand. Specifically, during each scheduling period, the supply-side power (including thermal power output, wind power output, photovoltaic power output, and pumped storage power) should satisfy an energy balance relationship with the pumping power of pumped storage, system load demand, and renewable energy reduction. This constraint ensures that the scheduling strategy 161 output by the joint clearing model 16 will not disrupt the system supply and demand balance.

[0083] System reserve constraint 142 is used to constrain the system's positive and negative spinning reserve capacity. As the proportion of renewable energy increases, the system needs to cope not only with positive reserve demand caused by increased load or decreased renewable energy output, but also with negative reserve demand caused by decreased load or increased renewable energy output. By setting system reserve constraint 142, the joint clearing model 16 can retain necessary adjustment margins while considering the lowest cost, avoiding the problem of optimal cost but insufficient safety margin.

[0084] The thermal power unit operating constraint 143 is used to constrain the unit's operating status, maximum technical output, minimum technical output, upward ramp rate, downward ramp rate, minimum continuous operating time, and minimum continuous downtime. Because thermal power units have start-up and shutdown inertia and ramp limitations, their output cannot be arbitrarily changed between adjacent time periods. Therefore, the thermal power unit operating constraint 143 is one of the important reasons for mixed-integer optimization modeling. By introducing 0-1 variables such as the unit's operating status, the unit's start-up, shutdown, start-up and shutdown costs, and deep peak-shaving selection can be expressed.

[0085] Renewable energy output constraint 144 is used to ensure that wind power output and photovoltaic output do not exceed their corresponding predicted or available output. This constraint allows the model to generate renewable energy reduction costs 155 when necessary, but cannot allow the actual output of new energy sources to exceed their physical available power. Since the environmental cost term 13 is introduced into the scheduling optimization objective 15, when the environmental cost of deep peak shaving is high, the model may choose to moderately curtail wind or solar power during certain periods to reduce the overall economic and environmental costs of the system.

[0086] Pumped storage operation constraint 145 is used to constrain the mutual exclusion of pumping and power generation states of a pumped storage power station within the same time period, and to constrain the water volume states of the upper and lower reservoirs and the energy conservation relationship. Pumped storage can absorb electricity during off-peak hours and generate electricity to support the system during peak hours, thus playing an important role in mitigating net load fluctuations 12 in the joint clearing model 16. By constraining the mutual exclusion of pumping and power generation states and the reservoir water volume boundary, it is possible to prevent the model from outputting a scheduling strategy that does not conform to the actual operating rules of the power station.

[0087] Example 6

[0088] Reference Figure 9 This embodiment illustrates the construction and solution method of the joint clearing model 16. When constructing the joint clearing model 16, the scheduling optimization objective 15 is set as the objective of minimizing the total system operating cost. The scheduling optimization objective 15 is synthesized from the basic power supply cost 151, the start-up and shutdown cost of thermal power units 152, the deep peak shaving compensation cost 153, the pumped storage operation cost 154, the renewable energy reduction cost 155, and the environmental cost item 13.

[0089] The basic electricity supply cost (151) reflects the fuel cost or market electricity purchase cost incurred by conventional generating units supplying electricity according to their output curves. The start-up and shutdown cost of thermal power units (152) reflects the economic cost of switching a unit from shutdown to startup or vice versa. The deep peak-shaving compensation cost (153) reflects the compensation required for thermal power units to deviate from their economic operating range after entering the deep peak-shaving zone. The pumped storage operation cost (154) reflects the cost of pumped hydro power generation, power generation losses, or pumped storage pricing. The renewable energy reduction cost (155) reflects the absorption losses caused by wind and solar curtailment. The environmental cost item (13) reflects the additional carbon emissions and pollutant treatment costs associated with deep peak-shaving.

[0090] In one specific implementation, the joint clearing model 16 can be in the form of mixed integer linear programming. Continuous variables in the model may include the output of each unit at each time period, the bid amount for deep peak shaving, the pumping power of pumped storage, the power generation of pumped storage, the actual output of renewable energy, and the amount of renewable energy reduction; discrete variables may include the unit start-up and shutdown status, unit start-up variables, unit shutdown variables, pumping status of pumped storage, power generation status of pumped storage, and the activation status of deep peak shaving.

[0091] Using scheduling optimization objective 15 as the objective function of joint clearing model 16, and scheduling constraint 14 as the constraint condition of joint clearing model 16, the solution can be obtained using CPLEX, Gurobi, CBC, or other mixed integer optimization solvers. The solution results can include scheduling strategy 161 that satisfies scheduling constraint 14 and the corresponding clearing cost 162. Scheduling strategy 161 can include the start-up and shutdown status of each thermal power unit, the output of each unit, the bid amount for peak shaving at each level, pumped storage pumping and power generation plans, and renewable energy reduction amounts, etc.

[0092] Let j be the set of electricity users, and let the load deviation of any electricity user j relative to its mean be... for: , In the formula: For load j in time period Load demand; It is a set of scheduling periods. Indicates time period The mean value of the internal load j.

[0093] set up A collection of fluctuating renewable energy sources, any renewable energy source The deviation of the output from its mean for: , In the formula: For renewable energy During the period of effort. Indicates time period Domestic renewable energy The mean.

[0094] System net load It can be represented as: , set up It is a set of peak-shaving demand entities, any peak-shaving demand entity During the period The sum of the values ​​contributing to the upward and downward fluctuations of net load is 0, that is... , In the formula: As the main body of peak demand During the period The deviation relative to its mean.

[0095] Under ideal conditions, when the output of all loads and fluctuating power sources is constant, the net load that conventional units need to meet is... During the period If the internal structure remains constant, it can be constructed as follows: , In the formula: This represents the ideal net load for the system. This represents the actual net load.

[0096] Considering that the main participants in paid peak shaving are thermal power units and pumped storage power stations, this embodiment establishes a scheduling optimization model that includes deep peak shaving and pumped storage. Its objective function is to minimize the overall "economic + environmental" coordinated cost of the system, i.e.: , in, This represents the total operating cost of the system. A collection of thermal power units; A collection of thermal power units participating in deep peak shaving; The set of thermal power units participating in basic peak shaving; A collection of peak-shaving gears; For the unit In the Quotations for peak shaving; For the unit During the period The The peak-shaving volume won in the bid for the peak-shaving segment; For the unit During the period The The environmental cost of deep peak shaving; Cost of thermal power units participating in basic peak shaving; A collection of pumped storage units; Indicates the unit During the period A 0-1 variable indicating whether it is enabled; a value of 1 indicates that it is enabled, otherwise it is 0. Indicates the unit During the period Startup costs; Indicates the unit During the period A 0-1 variable indicating whether the device is shut down; a value of 1 indicates shutdown, and a value of 0 indicates otherwise. Indicates the unit During the period Downtime costs; Indicates pumped storage unit During the period Pumping power; This indicates the price quoted for the pumping power of a pumped storage power station. For renewable energy The unit cost of electricity generation; For renewable energy During the period The actual power output; Indicates cost reduction per unit of energy; Represents renewable energy During the period The amount of reduction.

[0097] Relevant constraints include: System equilibrium constraints: , in, Indicates the unit During the period contribution; Indicates pumped storage unit During the period Power generation capacity.

[0098] System backup constraints: , , in, A 0-1 variable representing the operating status of the generator unit; a value of 1 indicates the generator unit is in good condition. During the period The device is powered on; a value of 0 indicates that it is powered off. and These represent the system during different time periods. Positive and negative backup requirements; and They represent the generating units. The maximum and minimum technical output.

[0099] Thermal power unit constraints: , , , , , , , , , For the unit During the period The minimum technical output for basic peak shaving; For the unit In the Maximum capacity of deep peak shaving; and The units Uphill and downhill climbing speed constraints; and The units Minimum continuous operation and downtime.

[0100] Renewable energy output constraints: , in, For renewable energy During the period The predicted output.

[0101] Constraints of pumped storage power stations: , , , , , , , , , In the formula: and Pumped storage units The minimum and maximum power generation capacity; and Pumped storage units The minimum and maximum pumping power; and These are pumped storage power stations and generating units. During the period The power generation status is a 0-1 variable, where 1 indicates that the system is in power generation status, and 0 otherwise. and These represent pumped storage power stations and generating units, respectively. During the period The pumping status is a 0-1 variable, where 1 indicates that the pumping state is in progress, and 0 otherwise. for Reservoir water storage at the end of the period and These are its upper and lower limits, respectively; and These are the ratios of average water volume to electricity conversion during pumping and power generation, respectively.

[0102] Example 7

[0103] Reference Figure 10 This embodiment illustrates the process of determining the characteristic function value 17. The characteristic function value 17 is used to describe the dynamic incremental cost caused by the net load fluctuation of any peak-shaving demand subject combination 221 after it is introduced into the system. In order to accurately isolate this dynamic incremental cost, a baseline alternative scenario 20 without peak-shaving demand is first constructed.

[0104] When constructing the baseline replacement scenario 20, the time-series power curves 101 of each peak-shaving demand subject 22 are replaced with their corresponding average lines. For example, the actual load curve of a user is replaced with the average load curve of that user within the scheduling period, and the actual output curves of wind power and photovoltaic power are replaced with their corresponding average output curves. After this processing, the system still has the same or similar total power volume, but the time-series fluctuations are weakened or eliminated. Therefore, the baseline replacement scenario 20 can serve as a reference scenario with no peak-shaving demand or low peak-shaving demand.

[0105] The joint clearing model 16 is invoked to optimize and clear the baseline replacement scenario 20, resulting in a baseline clearing cost 171. This baseline clearing cost 171 can be understood as the basic cost required for the system to meet energy balance and operational constraints when no significant time-series fluctuations occur among the various entities. Subsequently, for any combination of peak-shaving demand entities 221, the combined net load fluctuation 123 is generated based on the deviation of the time-series power curves 101 of each peak-shaving demand entity 22 within the combination from the corresponding mean line.

[0106] After superimposing the combined net load fluctuation 123 onto the baseline replacement scenario 20, a coalition peak-shaving scenario 21 is formed. Coalition peak-shaving scenario 21 only restores the fluctuation deviation corresponding to the combination of peak-shaving demand entities 221, without restoring the fluctuations of other entities not included in the combination. Through this construction method, the impact of different peak-shaving demand entity combinations 221 on system peak-shaving costs and environmental costs can be examined one by one. The joint clearing model 16 is then used to optimize and clear the coalition peak-shaving scenario 21, yielding the combined clearing cost 172.

[0107] Based on the difference between the combined clearing cost 172 and the benchmark clearing cost 171, the characteristic function value 17 corresponding to the peak-shaving demand subject combination 221 is determined.

[0108] By calculating the difference between the baseline replacement scenario 20 and the alliance peak shaving scenario 21, the dynamic incremental cost "caused by fluctuations" can be extracted more accurately compared to directly using the total clearing cost for allocation. This characteristic function value 17 provides the basis for subsequent Shapley value calculation and also makes the peak shaving ancillary service cost allocation result 19 interpretable: the cost allocated by each entity can be traced back to its impact on the characteristic function value 17 in different combined scenarios.

[0109] Example 8

[0110] Reference Figure 11 This embodiment illustrates the process of determining the marginal contribution 18. When determining the marginal contribution 18, the Shapley value rule is used to calculate the weighted average marginal contribution of the target entity 222 across different combinations 221 of peak-shaving demand entities. The target entity 222 can be any peak-shaving demand entity 22, such as a certain type of load entity, wind power entity, or photovoltaic entity.

[0111] For any target entity 222, first iterate through the peak-shaving demand entity combinations 221 that do not include target entity 222, then construct combinations that include target entity 222. Calculate the difference in characteristic function values ​​between the combinations containing target entity 222 and the combinations after removing target entity 222, respectively, to obtain the marginal cost increment 181 of target entity 222 under the corresponding combination. This marginal cost increment 181 reflects the impact of target entity 222 joining the system on dynamic incremental costs in a specific cooperation scenario.

[0112] Since the role of a particular entity may vary depending on the combined environment, its responsibility cannot be determined solely within a single combination. For example, a wind power entity may exacerbate peak-shaving pressure during off-peak periods, while a photovoltaic entity may reduce net load peaks during the day, thereby reducing the need for deep peak shaving by thermal power. The Shapley value rule, by traversing all possible combinations and weighting the marginal cost increment 181 under each combination, can obtain the average marginal contribution 18 of the target entity 222.

[0113] In one specific implementation, suppose there are n entities in the peak-shaving demand entity set, and S represents any peak-shaving demand entity combination 221 that does not include the target entity 222. Then, the marginal contribution 18 of the target entity 222 can be determined as follows: calculate the difference in characteristic function values ​​before and after the target entity 222 is added to combination S, multiply it by a weight related to the combination size, and finally sum over all combinations S. This weight is used to ensure that the average contribution under different addition orders is considered, thereby avoiding unfair allocation of contributions for a certain entity due to different calculation orders.

[0114] When the marginal contribution 18 is positive, it means that the average effect of the target entity 222 in all possible alliances is to increase the system's peak-shaving cost; when the marginal contribution 18 is negative, it means that the average effect of the target entity 222 is to reduce or offset the system's peak-shaving cost. Through this mechanism, this embodiment can identify the "positive adjustment value" that is difficult to reflect in traditional allocation methods. For example, photovoltaic output may have a certain match with the system's peak load during the day, and its fluctuations may actually reduce the probability of thermal power units entering the high-pollution deep adjustment range at certain times. In this case, the photovoltaic entity can exhibit a negative marginal contribution.

[0115] Specifically, it consists of four core steps:

[0116] The first step is data preprocessing. The input consists of 24-hour actual power curves for wind power, solar power, and load, which accurately depict the temporal pattern and physical scale of the actual output and demand of each entity. After cleaning and aligning the collected multi-source heterogeneous data, it is directly used as the input parameters for the underlying optimization calculation, laying the foundation for subsequent joint clearing modeling and cost accounting.

[0117] The second step is to construct and solve the alliance cost optimization model. Based on the preprocessed data, a mixed-integer linear programming model is adopted, with the goal of minimizing the system regulation cost. Taking into account conditions such as unit start-up and shutdown and output constraints, the minimum peak-shaving cost under all possible combinations of peak-shaving demand subjects (221) is calculated.

[0118] The third step is to determine the peak-shaving ancillary service cost to be allocated, which is 191. This involves precisely identifying the dynamic incremental costs caused by net load fluctuations, namely the deep peak-shaving and environmental externality cost pool. : , For peak-shaving market transactions and the costs of paid peak-shaving that need to be shared; This is a collection of deep peak shaving gears.

[0119] To reflect the mutual impact of different peak-shaving demand entities on system costs, this embodiment uses the Shapley value method to quantify the marginal contribution of each entity.

[0120] First, a baseline alternative scenario 20 with no peak-shaving demand is constructed, transforming the output curves of all peak-shaving demand entities into a stable average line. At this point, the system only needs to meet a constant power balance requirement, and the corresponding peak-shaving and environmental costs can be considered zero. Based on this, the system then targets the set of entities with peak-shaving needs. any subset to its fluctuation deviation The data is then overlaid onto the baseline scenario and the scheduling model is solved.

[0121] , in, Used for iterating over collections For all individual peak-shaving demand subjects, calculate the deviation of each subject from its own mean (i.e., ), and sum all these deviations to obtain the final set. Total fluctuation.

[0122] In any subset In the middle, the main body The marginal cost increment 181 is defined as: , In the formula, Indicates when the subject Fluctuations in the league The increase in peak-shaving costs resulting from incorporation into the system. When At that time, the main body Fluctuations exacerbate the system's peak-shaving pressure; when This indicates that its fluctuations help to smooth out the system load.

[0123] By traversing all possible alliance combinations, the principal is calculated. Weighted average marginal contribution : , Among them, weight It embodies the subset Includes main body At that time, the main body The percentage of influence within that subset. This is determined by iterating through all possible subsets. For subjects in different subsets The marginal contributions are weighted and integrated to obtain the final result. Able to reflect the subject The average contribution of the entity to the system's peak-shaving costs is also the key to the fairness of cost sharing achieved by the Shapley value, which fully considers the impact of the entity in all possible cooperation scenarios.

[0124] Finally, the fair allocation of peak-shaving costs is achieved based on the degree of triggering. To ensure a balanced distribution of revenue and expenditure during the allocation process, this embodiment adopts a proportional allocation model: , in, ; As the main body of peak demand The cost of paid peak shaving that should be shared.

[0125] Example 9

[0126] Reference Figure 12 This embodiment illustrates the process of generating the peak-shaving ancillary service cost allocation result 19. When generating the peak-shaving ancillary service cost allocation result 19, the peak-shaving ancillary service cost 191 to be allocated is determined based on the characteristic function value 17 corresponding to the peak-shaving demand subject combination 221, which includes all peak-shaving demand subjects 22. This peak-shaving ancillary service cost 191 to be allocated can be understood as a dynamic incremental cost pool caused by net load fluctuations under the combined effect of all peak-shaving demand subjects.

[0127] The allocation ratio 192 is determined based on the marginal contribution 18 of each peak-shaving demand entity 22. For entities with a positive marginal contribution 18, the allocation ratio 192 can be determined according to the proportion of their positive marginal contribution to the sum of all positive marginal contributions; for entities with a negative marginal contribution 18 or those representing a reduction effect, they can be treated as reduction entities and not directly bear the positive allocation cost, or their allocation cost 193 can be reduced to zero. The specific reduction method can be determined by market rules or settlement rules.

[0128] The peak-shaving ancillary service costs 191 to be allocated are distributed according to the allocation ratio 192, resulting in the allocated costs 193 corresponding to each peak-shaving demand entity 22. The allocated costs 193 can be output in monetary terms, or simultaneously output settlement auxiliary information such as allocation ratio, marginal contribution, and offset amount. Based on each allocated cost 193, a peak-shaving ancillary service cost allocation result 19 is generated. This result 19 can be used as input to the electricity market ancillary service settlement system, or as a basis for regulatory analysis, entity responsibility tracing, or low-carbon incentive evaluation.

[0129] When the marginal contribution 18 of a certain peak-shaving demand entity 22 indicates that it has a reducing effect on the system's peak-shaving costs, the allocated costs 193 of that peak-shaving demand entity 22 are reduced. Specifically, the allocated costs 193 of that entity can be set to zero, or its negative marginal contribution can be converted into a reduction contribution to reduce the allocated costs of other cost-bearing entities. Under the premise of ensuring that the allocated peak-shaving ancillary service costs 191 are ultimately balanced, this reduction process can reflect the positive regulatory value of that entity to the system.

[0130] In one simulation example, after allocating peak-shaving and environmental costs using the Shapley value method, it was found that the allocation ratio for different load-side entities is not simply equivalent to their electricity consumption percentage. Residential and service sector loads with large peak-to-valley differences may bear a higher proportion of peak-shaving costs even if their electricity consumption percentage is lower than that of the manufacturing sector; while photovoltaic (PV) loads, due to their daytime output which can mitigate some net load fluctuations, may exhibit a negative contribution. This example illustrates that the allocation method in this embodiment reflects load characteristics and the regulatory value of renewable energy, rather than mechanically allocating costs based on electricity volume.

[0131] Example 10

[0132] Referring to Tables 2 to 5, this embodiment provides a simulation verification method for a peak-shaving auxiliary service cooperative game cost-sharing method that takes into account environmental externalities.

[0133] Using a real power grid as a simulation object, the effectiveness of the peak-shaving cost-sharing mechanism described in this invention is verified. The unit composition of the power grid is shown in Table 2, with a total capacity of 38,400 MW, including a high proportion of wind power and photovoltaic installations. The basic peak-shaving benchmark for thermal power units is 50% of the rated capacity. When the basic peak-shaving of the system cannot meet the peak-shaving demand, pumped storage power stations and 36 thermal power units of 600 MW and 300 MW participate in competitive bidding for peak shaving.

[0134] Table 2: Power supply configuration.

[0135]

[0136] The example selects load, wind power, and photovoltaic as the main peak-shaving demands, and sets the dispatch cycle to 24 hours. The daily system load curve and the output characteristics of wind power and photovoltaic are as follows: Figure 15 As shown. The minimum configuration standard for system load reserve is set at 5% of the total system load, and the additional reserve capacity configured for renewable energy output fluctuations is 5% of its predicted output. The energy reduction unit price is calculated at 500 yuan / (MW·h). The constructed scheduling optimization model belongs to the category of mixed integer optimization. Based on the optimization orientation of minimizing the total system cost, the model uses the CPLEX solver on the GAMS platform to complete the clearing calculation for the 24-hour scheduling cycle. The per-unit value curve of peak demand and the load forecast curve are referenced. Figure 15 .

[0137] Based on the optimization objective of minimizing the total system cost, the clearing results show that the total peak-shaving cost is RMB 2.6623 million, and the total paid peak-shaving electricity is 9,015 MW·h. Considering the environmental externality costs, the system proactively implemented wind curtailment measures during specific periods (periods 2 to 5 and period 24), with a cumulative wind curtailment of 2,924.86 MW·h. This phenomenon reflects that in the low-carbon dispatch model proposed in this embodiment, when the sum of the environmental pollution cost and unit regulation cost generated by deep peak shaving exceeds the energy curtailment penalty, the system will spontaneously optimize overall benefits through moderate wind curtailment, rather than blindly pursuing full absorption. Combined with the deep peak-shaving statistical data analysis in Table 3, the 600MW unit is the core entity undertaking peak-shaving demand. During periods of low load and drastic fluctuations in net load, the large-capacity unit demonstrated stronger regulation support capabilities.

[0138] The demand for deep peak shaving is highly concentrated on the timeline, mainly distributed during the early morning off-peak period (periods 1-6) and at night (periods 23-24). The demand for deep peak shaving peaks in period 3 reaches its peak, with a total awarded capacity of 3,030 MW, all undertaken by 600MW units. Because the carbon emissions and pollutant emissions per unit of electricity corresponding to the fourth tier of deep peak shaving are much higher than those in the basic peak shaving stage, although the model utilized extremely deep peak shaving capacity in period 3, a total of 2126.33MW of wind curtailment space still formed from periods 2 to 5. This demonstrates that under an optimization mechanism including environmental costs, the system tends to avoid the surge in environmental externalities caused by units being in extremely deep peak shaving state for extended periods through the coordinated operation of unit peak shaving and moderate energy curtailment. The peak shaving clearing results for each period are as follows: Figure 16 and Figure 17 As shown, refer to Figure 16 and Figure 17 In the graph, the horizontal axis represents the scheduling period, the vertical axis represents the generator unit number, and the vertical axis represents the peak-shaving level. A vertical axis value of 0 indicates that the corresponding generator unit is shut down or not called during the corresponding scheduling period; a vertical axis value of 1 indicates that the corresponding generator unit is in basic peak-shaving mode; vertical axis values ​​of 2, 3, and 4 indicate that the corresponding generator unit has entered a deeper peak-shaving level, respectively. The different colored curves in the graph are used to distinguish the peak-shaving level trajectories corresponding to different generator unit numbers in a three-dimensional coordinate system, making the level changes of different generator units in different scheduling periods clearer. Different colors do not inherently correspond to different peak-shaving demand entities, different environmental cost items, or different calculation models; the specific meaning of peak-shaving is determined by the corresponding position of the curve on the horizontal, vertical, and horizontal axes. Therefore, it is possible to... Figure 16 and Figure 17 This visually demonstrates the differences in the dispatching of 300MW and 600MW units during various scheduling periods, as well as the degree of responsibility undertaken by units of different capacity levels during the deep peak shaving and clearing process.

[0139] Table 3: Overall depth peak shaving statistics.

[0140]

[0141] The system comprises five peak-shaving demand entities (members), denoted as M1, M2, M3, N1, and N2 for residential load, manufacturing load, service load, wind power, and solar power, respectively. Table 4 details the distribution of peak-shaving costs and environmental costs under different peak-shaving demand entity alliances. The set of all peak-shaving demand entities can be represented as... Before quantifying the costs of each subset, a globally unique scenario without peak-shaving demand substitution is first constructed as a reference benchmark, and the load and renewable energy output curves are converted into mean lines without fluctuations. Based on this, the corresponding fluctuation deviations are superimposed on each of the 32 subsets, and a scheduling optimization model including environmental costs is solved to calculate the system peak-shaving cost and environmental cost for each subset. The example results show that the system's basic peak-shaving capacity can offset some of the adjustment demand caused by the main fluctuations, resulting in 22 subsets with zero peak-shaving costs; the remaining 10 are typical subsets with non-zero peak-shaving costs.

[0142] Based on the calculation results in Table 4, the system peak-shaving cost exhibits significant nonlinear combination characteristics under different entity alliance scenarios. In scenarios with smaller alliance sizes or lower load fluctuations, such as... or The collection shows that peak-shaving costs remain at a low level and environmental costs are all zero. This reflects that under the guidance of optimal total system cost, net load demand is mainly covered by the basic technical output of conventional units, without triggering a high-cost deep peak-shaving state.

[0143] However, with the superposition of highly volatile entities or multiple types of loads, the system is forced to utilize deep peak-shaving resources, leading to a surge in costs. Taking the alliance as an example, its peak-shaving cost reached 3.8137 million yuan, and the corresponding environmental cost also emerged and increased to 136,400 yuan. The introduction of environmental costs objectively quantifies the additional carbon emissions and pollutant costs incurred by thermal power units operating under low-load conditions. When wind power... Access After the scenario, the peak-shaving cost and environmental cost further increased to RMB 4.4109 million and RMB 163,000 respectively, verifying that wind power output has strong anti-peak-shaving characteristics, which increased the economic and environmental burden of the system.

[0144] In stark contrast, photovoltaics The grid connection demonstrated significant positive regulatory value. After photovoltaic (PV) entities joined the alliance, peak-shaving costs plummeted from 3.8137 million yuan to 366,600 yuan, and environmental costs were reduced to 11,100 yuan. This result demonstrates that PV output effectively alleviated the system's peak-valley regulation pressure and reduced the frequency of thermal power plants entering high-pollution peak-shaving zones. Ultimately, in the full-member alliance scenario, the total system peak-shaving cost was fixed at 2.6623 million yuan, and the environmental cost at 85,900 yuan. This cost quantification method, based on feedback from the joint optimization of energy and ancillary services, can more accurately capture the marginal contribution of each demand entity, laying a solid physical and economic foundation for subsequent equitable allocation, including environmental responsibility, using Shapley values.

[0145] Table 4: Peak shaving costs and environmental costs (ten thousand yuan) for each set.

[0146]

[0147] This embodiment uses the Shapley value method to allocate peak shaving and environmental costs, and scientifically defines the weighted average marginal contribution of each entity to the total system cost.

[0148] On the load side, residential load Service sector load Although their total electricity consumption accounts for only 28% and 18% respectively, due to their large peak-to-valley differences and drastic fluctuations, they respectively share 32.6% and 32.27% of the peak-shaving costs. In contrast, the manufacturing load... Although it consumed 54% of the electricity, its load curve was relatively stable, resulting in an environmental cost-sharing ratio of only 24.02%. The environmental cost-sharing ratio also showed a similar trend, demonstrating that load characteristics, rather than the scale of electricity consumption, are the primary cause of peak-shaving environmental responsibility. On the power generation side, wind power... It exhibits significant "anti-peak-shaving" characteristics, resulting in a 13.5% contribution to peak-shaving costs and a 13.74% contribution to the environment, ultimately bearing 11.11% of the paid peak-shaving fees. Meanwhile, photovoltaic... The Shapley value contribution is negative in both peak shaving and environmental dimensions (peak shaving cost contribution of -571,700 yuan and environmental contribution of -23,300 yuan), reflecting that photovoltaics can effectively smooth net load fluctuations during the day and contribute significant "ecological value". Although the photovoltaic cost-sharing is set to 0 according to market rules, its negative contribution substantially reduces the payment pressure on other members in the calculation process.

[0149] The above-mentioned peak-shaving cost contribution and environmental contribution value are the specific manifestations of marginal contribution 18 in the dimensions of peak-shaving cost and environmental cost.

[0150] Table 5: Allocation of peak-shaving costs and environmental costs (ten thousand yuan).

[0151]

[0152] In the process of building a new power system, how to streamline the economic relationship of peak-shaving ancillary services through a scientific cost-sharing mechanism is crucial to ensuring the safe and stable operation of the power system. This invention systematically explores the method of quantifying the whole system cost and the cost-sharing mechanism based on Shapley values ​​in the deep peak-shaving process by constructing a joint clearing model of electricity and peak-shaving ancillary services that takes into account environmental externalities, namely, Joint Clearing Model 16. Experimental results show that internalizing the costs of carbon emissions and pollutant reduction into the peak-shaving clearing model can accurately characterize the environmental burden caused by the efficiency decline of thermal power units under low-load conditions; and the cost sharing achieved through game theory not only reflects the marginal contribution of each entity to the system's peak-shaving demand, but also reveals the positive regulatory value of grid connection of new energy sources (such as photovoltaics) in alleviating system regulation pressure and reducing the environmental cost of the whole system.

[0153] Example 11

[0154] Reference Figure 13 This embodiment provides a peak-shaving auxiliary service cooperative game cost-sharing system that takes into account environmental externalities. The system includes a data processing unit 100, a joint clearing unit 200, and a cost-sharing unit 300. Each unit can be configured in the same server, or it can be deployed separately in a scheduling data platform, a joint clearing platform, and a market settlement platform, and connected through a communication interface, a database, or a message queue.

[0155] The data processing unit 100 acquires system time-series data 10 and unit operation emission data 11, and generates net load fluctuation 12 and environmental cost item 13 based on the system time-series data 10 and the unit operation emission data 11, respectively. The data processing unit 100 may include a data acquisition interface, a data cleaning module, a time scale alignment module, a mean baseline generation module, a deviation calculation module, and an environmental cost calculation module. The net load fluctuation 12 and environmental cost item 13 output by the data processing unit 100 are sent to the joint clearing unit 200.

[0156] The joint clearing unit 200 is connected to the data processing unit 100 and is used to receive net load fluctuations 12 and environmental cost items 13. Based on scheduling constraints 14 and the scheduling optimization objective 15 that incorporates environmental cost items 13, it constructs a joint clearing model 16 to determine the characteristic function value 17. The joint clearing unit 200 may include an objective function generation module, a constraint generation module, a scenario construction module, a mixed-integer optimization solution module, and a clearing result processing module. The scenario construction module can generate a baseline substitution scenario 20 and a coalition peak-shaving scenario 21. After the optimization solution module solves the different scenarios separately, the clearing result processing module obtains the characteristic function value 17 based on the clearing cost difference.

[0157] The cost allocation unit 300 is connected to the joint clearing unit 200 and is used to receive the characteristic function value 17, determine the marginal contribution 18 of each peak-shaving demand entity 22 based on the characteristic function value 17, and generate the peak-shaving ancillary service cost allocation result 19 based on the marginal contribution 18. The cost allocation unit 300 may include a characteristic function storage module, a marginal cost increment calculation module, a Shapley weighted calculation module, an allocation ratio calculation module, a reduction processing module, and a result output module. The result output module can output the allocated cost 193, allocation ratio 192, marginal contribution 18, and an indicator of whether there is a reduction processing for each peak-shaving demand entity 22.

[0158] Finally, it should be noted that the methods and systems described in detail above are merely embodiments, and those skilled in the art can modify these embodiments in different ways as long as they do not depart from the scope of the present invention.

Claims

1. A method for cost allocation in a cooperative game theory approach to peak-shaving ancillary services, taking into account environmental externalities, characterized in that: include, Acquire system time-series data and unit operation emission data, determine the net load fluctuation of the main peak-shaving demand based on the system time-series data, and generate an environmental cost item characterizing the environmental externality of deep peak-shaving of thermal power units based on the unit operation emission data. Define scheduling constraints, incorporate the environmental cost item into the joint clearing scheduling optimization objective, and use the scheduling optimization objective as the objective function and the scheduling constraints as the constraint conditions to construct a joint clearing model that takes environmental costs into account using a mixed integer optimization modeling method; The scheduling strategy is optimized through the joint clearing model. The dynamic incremental cost caused by the net load fluctuation is separated as the characteristic function value. The marginal contribution of each peak-shaving demand subject is determined according to the characteristic function value, and the peak-shaving auxiliary service cost allocation result is generated based on the marginal contribution.

2. The peak-shaving ancillary service cooperative game cost-sharing method considering environmental externalities as described in claim 1, characterized in that: The system time-series data includes the time-series power curves of the main peak-shaving demand entities, and the time-series power curves include the actual output curves of wind power, photovoltaic power, and actual user load. The unit's operating emission data includes environmental price parameters and unit operating condition parameters. The environmental price parameters include carbon price, sulfur dioxide environmental treatment unit price, and nitrogen oxide environmental treatment unit price. The unit operating condition parameters include unit output, unit capacity level, deep peak shaving level, minimum output for deep peak shaving without oil injection, and coal consumption increment.

3. The peak-shaving ancillary service cooperative game cost-sharing method considering environmental externalities as described in claim 2, characterized in that: When determining the net load fluctuation, the actual user load curve, the actual wind power output curve, and the actual photovoltaic output curve are unified to the same time scale. The load deviation component is obtained based on the deviation of the user's actual load curve from the corresponding load average baseline. Based on the deviation of the actual wind power output curve and the actual photovoltaic power output curve from the corresponding new energy power output average baseline, the new energy power output deviation component is obtained; The load deviation component is taken as the positive component that increases the net load of the system, and the new energy output deviation component is taken as the negative component that reduces the net load of the system. The load deviation component and the new energy output deviation component are combined to obtain the net load fluctuation.

4. The peak-shaving ancillary service cooperative game cost-sharing method considering environmental externalities as described in claim 2 or 3, characterized in that: When generating the environmental cost item, the peak-shaving operation range of the thermal power unit is determined based on the unit operating parameters, and the emission increment parameters corresponding to the peak-shaving operation range are obtained. The emission discount price is determined based on the aforementioned environmental price parameters; Using the paid peak-shaving benchmark as the calculation basis, the additional carbon emission costs caused by the increase in coal combustion and the pollution gas treatment costs caused by the increase in oil production are calculated based on the emission increment parameters and the emission conversion price. The additional carbon emission costs are combined with the pollution gas treatment costs to form the environmental cost item.

5. The peak-shaving ancillary service cooperative game cost-sharing method considering environmental externalities as described in any one of claims 1 to 3, characterized in that: The scheduling constraints include system balance constraints, system reserve constraints, thermal power unit operation constraints, renewable energy output constraints, and pumped storage operation constraints. The system balance constraint is used to constrain the balance relationship between thermal power output, renewable energy output, pumped storage power and system load demand; The system reserve constraint is used to constrain the system's positive spin-off reserve capacity and negative spin-off reserve capacity; The thermal power unit operation constraints are used to constrain the unit's operating status, maximum technical output, minimum technical output, upward ramp rate, downward ramp rate, minimum continuous operating time, and minimum continuous downtime. The renewable energy output constraint is used to ensure that wind power output and photovoltaic output do not exceed the corresponding predicted output. The pumped storage operation constraints are used to constrain the pumping and power generation states of a pumped storage power station to be mutually exclusive during the same period, and to constrain the water volume states of the upper and lower reservoirs and the energy conservation relationship.

6. The peak-shaving ancillary service cooperative game cost-sharing method considering environmental externalities as described in claim 5, characterized in that: When constructing the joint clearing model, the scheduling optimization objective is set as minimizing the total system operating cost; The scheduling optimization objective is composed of the basic electricity supply cost, thermal power unit start-up and shutdown cost, deep peak shaving compensation cost, pumped storage operation cost, renewable energy reduction cost, and the environmental cost item. The scheduling optimization objective is used as the objective function of the joint clearing model, and the scheduling constraints are used as the constraints of the joint clearing model. The joint clearing model is optimized to obtain a scheduling strategy that satisfies the scheduling constraints.

7. The peak-shaving ancillary service cooperative game cost-sharing method considering environmental externalities as described in claim 3, characterized in that: When determining the characteristic function value, the time-series power curves of each of the peak-shaving demand subjects are replaced with the corresponding mean line to form a benchmark alternative scenario with no peak-shaving demand. The joint clearing model is invoked to optimize the clearing of the benchmark alternative scenario, thereby obtaining the benchmark clearing cost; For any combination of peak-shaving demand entities, the combined net load fluctuation is generated based on the deviation of the time-series power curves of each peak-shaving demand entity within the combination from the corresponding mean line. The combined net load fluctuations are superimposed on the baseline alternative scenario to form a coalition peak shaving scenario; The joint clearing model is invoked to optimize the clearing of the alliance peak shaving scenario, and the combined clearing cost is obtained; The characteristic function value corresponding to the peak-shaving demand subject combination is determined based on the difference between the combined clearing cost and the benchmark clearing cost.

8. The peak-shaving ancillary service cooperative game cost-sharing method considering environmental externalities as described in claim 7, characterized in that: When determining the marginal contribution, the Shapley value rule is used to calculate the weighted average marginal contribution of the target entity in different combinations of peak-shaving demand entities; Specifically, for any target entity, the difference in characteristic function values ​​between the combination of peak-shaving demand entities that includes the target entity and the combination of peak-shaving demand entities that excludes the target entity is calculated to obtain the marginal cost increment of the target entity under the corresponding combination. The marginal cost increment of the target entity under different combinations of peak-shaving demand entities is weighted to obtain the marginal contribution of the target entity.

9. The peak-shaving ancillary service cooperative game cost-sharing method considering environmental externalities as described in claim 7 or 8, characterized in that: When generating the peak shaving ancillary service cost allocation result, the peak shaving ancillary service cost to be allocated is determined based on the characteristic function value corresponding to the combination of peak shaving demand subjects that includes all the peak shaving demand subjects. The corresponding sharing ratio is determined based on the marginal contribution of each of the aforementioned peak-shaving demand entities; The cost of the peak-shaving auxiliary service to be allocated is allocated according to the allocation ratio to obtain the allocated cost corresponding to each of the peak-shaving demand entities; The peak-shaving ancillary service cost allocation result is generated based on each of the aforementioned allocated costs; Specifically, when the marginal contribution of the peak-shaving demand subject indicates that it has a reducing effect on the system's peak-shaving cost, the allocated cost of the peak-shaving demand subject is reduced.

10. A peak-shaving auxiliary service cooperative game cost-sharing system that takes into account environmental externalities, characterized in that, include: The data processing unit is used to acquire system time-series data and unit operation emission data, and generate net load fluctuation and environmental cost items based on the system time-series data and the unit operation emission data, respectively. A joint clearing unit, connected to the data processing unit, is used to receive the net load fluctuation and the environmental cost item, and to construct a joint clearing model based on scheduling constraints and the scheduling optimization objective that incorporates the environmental cost item, so as to determine the characteristic function value through the joint clearing model; The cost allocation unit, connected to the joint clearing unit, is used to receive the characteristic function value, determine the marginal contribution of each peak-shaving demand subject based on the characteristic function value, and generate the peak-shaving auxiliary service cost allocation result based on the marginal contribution.

Images