CN-121480841-B - Energy storage power station business mode collaborative optimization method and system based on multi-objective planning

CN121480841BCN 121480841 BCN121480841 BCN 121480841BCN-121480841-B

Abstract

The invention provides a multi-objective planning-based energy storage power station business mode collaborative optimization method and a multi-objective planning-based energy storage power station business mode collaborative optimization system, which relate to the technical field of energy storage power station operation and comprise the steps of performing space-time dimension decoupling on energy storage power station operation data to obtain a business mode feature set; the method comprises the steps of establishing a multi-objective planning constraint system, obtaining a business mode combination scheme based on a pareto front search strategy, generating a scheduling execution instruction according to the business mode combination scheme, identifying business modes with conflicts and carrying out capacity allocation adjustment. The invention realizes the dynamic collaborative optimization of multiple business modes of the energy storage power station, and improves the economic benefit and the operation safety.

Inventors

PAN YINGCHAO
DUAN XIAOHAN
XIAN CE

Assignees

北京如实新能源有限公司

Dates

Publication Date: 20260508
Application Date: 20251106

Claims (9)

1. The multi-objective planning-based energy storage power station business model collaborative optimization method is characterized by comprising the following steps of: Performing space-time dimension decoupling on operation data of the energy storage power station to obtain a commercial mode feature set, combining the commercial mode feature set with an economic optimization target and a risk constraint optimization target, and establishing a multi-target planning constraint system, wherein the method comprises the following steps: the energy storage power station operation data comprise charge and discharge power time sequence data, charge state time sequence data and load demand time sequence data; performing time scale decomposition on the charge-discharge power time sequence data, and extracting peak-valley period characteristics and periodic characteristics; performing space state mapping on the state of charge time sequence data, identifying a state transition path, and extracting capacity utilization depth characteristics and state duration characteristics corresponding to the state transition path; Calculating the coupling relation between the peak-valley period characteristics and the capacity utilization depth characteristics, generating a time-shifting schedulable characteristic vector, calculating the coupling relation between the periodic characteristics and the state duration characteristics, generating an auxiliary service schedulable characteristic vector, calculating the matching relation between the load demand time sequence data and the state transition path, and generating a demand management schedulable characteristic vector; Forming the business model feature set from the time-shifted schedulable feature vector, the auxiliary service schedulable feature vector and the demand management schedulable feature vector; The economic optimization target quantifies the gain sum of all business modes, the risk constraint optimization target quantifies the coupling conflict degree between business modes, the economic optimization target and the risk constraint optimization target are taken as dual optimization targets, the business mode feature set is taken as a decision variable, and the multi-target planning constraint system is established; performing non-dominant solution set iterative solution on the multi-objective planning constraint system based on a pareto front search strategy, dynamically adjusting weight coefficients of schedulable feature vectors in the business model feature set according to the risk constraint optimization target in the solution process, and obtaining an optimal solution set of the economic optimization target under the condition of meeting the risk constraint optimization target to obtain a business model combination scheme; generating a scheduling execution instruction according to the time sequence distribution proportion of each business mode in the business mode combination scheme and combining capacity state data of the energy storage power station, and executing the scheduling execution instruction; In the process of executing the instruction, identifying the business modes with time sequence overlapping and capacity overrun in the business mode combination scheme, establishing time sequence priority ordering by calculating the occupied demand quantity and the release time of each business mode on the capacity state data, carrying out capacity allocation adjustment on the business modes with conflicts according to the time sequence priority ordering, and feeding back the adjusted capacity allocation result to the schedulable feature vectors corresponding to the business mode feature sets, so as to realize dynamic coordination among the business modes.
2. The method of claim 1, wherein iteratively solving the multi-objective planning constraint system for a non-dominant solution set based on a pareto front search strategy, wherein dynamically adjusting the weight coefficients of each schedulable feature vector in the business model feature set according to the risk constraint optimization objective during the solving comprises: Generating an initial solution set in a decision space corresponding to the business model feature set; Calculating a coupling conflict degree value of the benefit value of the economic optimization target and the risk constraint optimization target, constructing a double-target fitness space, and identifying non-dominant solutions of the initial solution set in the double-target fitness space and storing the non-dominant solutions to a pareto front edge set; Based on the coupling conflict degree value corresponding to each non-dominant solution in the pareto front edge set, calculating conflict contribution degree of the business model feature set on time sequence occupation and capacity occupation, and carrying out self-adaptive adjustment on weight coefficients among a time-shifting schedulable feature vector, an auxiliary service schedulable feature vector and a demand management schedulable feature vector according to the conflict contribution degree and a non-dominant solution with the maximum economic optimization target profit value to generate a candidate solution set; and comparing the candidate solution set with the pareto front set in a non-dominant relation, and iteratively updating the pareto front set based on the result of the non-dominant relation comparison.
3. The method of claim 1, wherein obtaining an optimal solution set for the economic optimization objective under conditions that satisfy the risk constraint optimization objective, the obtaining a business model combination solution comprises: Calculating the total capacity occupation amount corresponding to each non-dominant solution in the pareto front edge set, and identifying a period when the total capacity occupation amount exceeds the rated capacity of the energy storage power station as a capacity overrun risk index; Calculating the number of overlapping time periods of the business model feature set in a time sequence dimension, and taking the number of overlapping time periods as a time sequence conflict risk index; Judging whether the capacity overrun risk index and the time sequence conflict risk index corresponding to each non-dominant solution in the pareto front edge set meet the threshold condition of the risk constraint optimization target, selecting the non-dominant solution with the largest economic optimization target profit value from the pareto front edge set meeting the threshold condition, and taking the weight coefficient of each schedulable feature vector corresponding to the non-dominant solution as an optimal solution set; Based on the optimal solution set, the time sequence distribution proportion of the electric energy time shift business mode, the auxiliary service response business mode and the demand management business mode is respectively adjusted in a planning period, and the business mode combination scheme is generated by combining peak-valley period characteristics, periodic characteristics and matching relations in the demand management schedulable characteristic vector.
4. The method of claim 1, wherein generating scheduling execution instructions in conjunction with capacity state data of an energy storage power station and executing according to a time-sequential allocation ratio of each business mode in the business mode combination scheme comprises: Acquiring capacity state data of an energy storage power station, wherein the capacity state data comprises a current state of charge value, available charge capacity and available discharge capacity; dividing a planning period into a plurality of scheduling time slots, determining the time duration duty ratio of each scheduling time slot to the electric energy time shift business mode, the auxiliary service response business mode and the demand management business mode, and generating a time sequence resource allocation matrix; For each scheduling time slot, judging a charging interval, a discharging interval or a standby interval where an energy storage power station is located according to the current state of charge value, and distributing the available charging capacity and the available discharging capacity to generate a capacity resource distribution matrix; Calculating capacity shares and duration allocated to corresponding business modes based on the time sequence resource allocation matrix and the capacity resource allocation matrix to obtain charge and discharge power instruction values, and generating the scheduling execution instruction by combining business mode execution sequences of each scheduling time slot; And issuing the scheduling execution instruction to a power control unit of the energy storage power station, wherein the power control unit executes charging operation or discharging operation.
5. The method of claim 1, wherein identifying business patterns in the business pattern combination scheme that have time sequence overlaps and capacity overruns, and establishing a time sequence prioritization by calculating an occupancy demand and a release moment of each business pattern for the capacity state data comprises: counting the overlapping time length of the actual execution time periods among all business modes, and marking the business mode combination with the overlapping time length larger than zero as a time sequence overlapping business mode combination; For each business mode in the time sequence overlapped business mode combination, based on the charge and discharge power instruction value corresponding to each business mode and the duration of the actual execution period, obtaining and summing the capacity occupation demand of each business mode on the capacity state data, and if the summation result exceeds the available charge capacity or the available discharge capacity of the energy storage power station, marking the time sequence overlapped business mode combination as a capacity overrun business mode combination; Calculating the release time of each business mode to the capacity of the energy storage power station according to each business mode in the capacity overrun business mode combination, wherein the release time is the end time of the actual execution period of each business mode, and calculating the capacity occupation duration of each business mode based on the release time; And establishing a time sequence priority ordering rule based on the capacity occupation demand and the capacity occupation duration, giving the highest time sequence priority to the business mode with the shortest capacity occupation duration and the smallest capacity occupation demand, and performing descending order to generate the time sequence priority ordering.
6. The method of claim 1, wherein performing capacity allocation adjustment on the business models with conflicts according to the time sequence priority ordering, and feeding back the adjusted capacity allocation result to the schedulable feature vectors corresponding to the feature sets of the business models, and wherein implementing dynamic collaboration between the business models comprises: According to the sequence from high to low of the time sequence priority, the available charge capacity or the available discharge capacity of the energy storage power station is distributed for each business mode in sequence, and the full distribution is carried out according to the capacity occupation demand of the business mode; When the residual capacity of the energy storage power station is smaller than the capacity occupation demand of the current business mode, distributing the residual capacity to the current business mode, and calculating the capacity shortage of the current business mode; Generating an adjusted capacity allocation value for each business model; Feeding back the adjusted capacity allocation value to the business model feature set, and based on the updated capacity allocation parameters in the business model feature set, recalculating actual capacity occupation total amounts of the time-shifting schedulable feature vector, the auxiliary service schedulable feature vector and the demand management schedulable feature vector in the current scheduling time slot, and judging whether the actual capacity occupation total amounts meet capacity constraint conditions of an energy storage power station or not; Generating a compensation instruction for the business mode with the capacity deficiency and executing in the current scheduling time slot to compensate the capacity deficiency; And updating the dispatching execution instruction to realize dynamic collaboration among business modes.
7. A multi-objective planning-based energy storage power station business model collaborative optimization system for implementing the method of any of claims 1-6, comprising: The system comprises a first unit, a second unit and a third unit, wherein the first unit is used for performing space-time dimensional decoupling on operation data of an energy storage power station to obtain a commercial mode feature set, combining an economic optimization target with a risk constraint optimization target and establishing a multi-target planning constraint system; The second unit is used for carrying out non-dominant solution set iterative solution on the multi-objective planning constraint system based on a pareto front search strategy, dynamically adjusting the weight coefficient of each schedulable feature vector in the commercial mode feature set according to the risk constraint optimization target in the solution process, and obtaining the optimal solution set of the economic optimization target under the condition of meeting the risk constraint optimization target to obtain a commercial mode combination scheme; the third unit is used for generating and executing a scheduling execution instruction according to the time sequence distribution proportion of each business mode in the business mode combination scheme and combining capacity state data of the energy storage power station; and a fourth unit, configured to identify, in the instruction execution process, a business mode in which there is a time sequence overlap and a capacity overrun in the business mode combination scheme, establish a time sequence priority ranking by calculating an occupation demand and a release time of each business mode on the capacity status data, perform capacity allocation adjustment on the business mode in which there is a conflict according to the time sequence priority ranking, and feed back an adjusted capacity allocation result to a schedulable feature vector corresponding to the feature set of the business mode, so as to implement dynamic coordination between the business modes.
8. An electronic device, comprising: A processor; A memory for storing processor-executable instructions; Wherein the processor is configured to invoke the instructions stored in the memory to perform the method of any of claims 1 to 6.
9. A computer readable storage medium having stored thereon computer program instructions, which when executed by a processor, implement the method of any of claims 1 to 6.

Description

Energy storage power station business mode collaborative optimization method and system based on multi-objective planning Technical Field The invention relates to the technical field of operation of energy storage power stations, in particular to a business model collaborative optimization method and a business model collaborative optimization system of an energy storage power station based on multi-objective planning. Background Along with the large-scale grid connection of renewable energy sources and the transformation of power system structures, the energy storage power station plays an increasingly important role in the aspects of peak regulation, frequency modulation, spare capacity, demand response and the like as an important power grid regulation resource. The energy storage power station can participate in various business modes simultaneously, including an electric auxiliary service market, an electric energy market, a capacity market and the like, so as to improve the economic benefit and the social value. The business mode optimization of the energy storage power station is a key problem in the current power system operation and energy conversion process, and has important significance in promoting new energy consumption and guaranteeing the stable operation of the power grid. The traditional optimization method cannot effectively decouple the space-time dimension characteristics of the business modes of the energy storage power station, and is difficult to accurately capture the interaction among different business modes and the differentiated requirements of the energy storage resources, so that the difference exists between the optimization result and the actual running condition. The existing business model optimization method is usually guided by a single economic target, lacks dynamic adjustment capability for risk constraint, cannot realize flexible balance between benefits and risks, and particularly cannot guarantee stable operation and long-term benefits of the energy storage power station under the condition of severe fluctuation of market environment. When the energy storage power station operates in multiple business modes at the same time, capacity competition and time sequence conflict often occur, the prior art lacks an effective real-time collaborative optimization mechanism, the resource allocation of each business mode cannot be dynamically adjusted according to the actual operation state, seamless connection and global optimization between business modes are difficult to realize, and the utilization efficiency of energy storage resources is reduced. Disclosure of Invention The embodiment of the invention provides a multi-objective planning-based energy storage power station business model collaborative optimization method and system, which can solve the problems in the prior art. In a first aspect of an embodiment of the present invention, there is provided a multi-objective planning-based collaborative optimization method for business models of an energy storage power station, including: Performing space-time dimension decoupling on operation data of an energy storage power station to obtain a commercial mode feature set, combining the commercial mode feature set with an economic optimization target and a risk constraint optimization target, and establishing a multi-target planning constraint system; performing non-dominant solution set iterative solution on the multi-objective planning constraint system based on a pareto front search strategy, dynamically adjusting weight coefficients of schedulable feature vectors in the business model feature set according to the risk constraint optimization target in the solution process, and obtaining an optimal solution set of the economic optimization target under the condition of meeting the risk constraint optimization target to obtain a business model combination scheme; generating a scheduling execution instruction according to the time sequence distribution proportion of each business mode in the business mode combination scheme and combining capacity state data of the energy storage power station, and executing the scheduling execution instruction; In the process of executing the instruction, identifying the business modes with time sequence overlapping and capacity overrun in the business mode combination scheme, establishing time sequence priority ordering by calculating the occupied demand quantity and the release time of each business mode on the capacity state data, carrying out capacity allocation adjustment on the business modes with conflicts according to the time sequence priority ordering, and feeding back the adjusted capacity allocation result to the schedulable feature vectors corresponding to the business mode feature sets, so as to realize dynamic coordination among the business modes. Performing space-time dimension decoupling on operation data of the energy storage power station to obtain a commercial mode feat