34results about How to "Reduce light rejection" patented technology

The invention discloses an electric heating energy storage poly-generation system comprising a heat pump system, a heat and cold storage system, a heat supply system, a refrigeration system and a power generation system. The heat pump system comprises a motor, a compressor, a heat storage heat exchanger and a cold storage heat exchanger. The heat and cold storage system comprises a high-temperature storage tank, a normal-temperature storage tank and a low-temperature storage tank. By means of the system, the renewable energy source electricity or electricity generated when the electrical network load is at the trough can be stored in the form of heat and coldness, and meanwhile, the stored heat or coldness can be used for conducting heating, refrigeration and power generation.

The invention discloses a coordinated scheduling method for an electricity-hydrogen multi-energy complementary DC microgrid, and the method is established based on the electricity-hydrogen multi-energy complementary DC microgrid, and comprises the following steps: initializing each system; performing the day-ahead prediction of the output of a renewable energy power generation system and the localload demand; performing day-ahead economic dispatching based on the predicted power, the electric energy storage system reserve SOC and the hydrogen energy system reserve SOHC; predicting photovoltaic, fan output and local load requirements in real time in an ultra-short period; and carrying out real-time rolling optimization on the day-ahead economic dispatching result according to the ultra-short-term prediction result. According to the invention, through day-ahead and online coordinated scheduling, the normal operation of the electricity-hydrogen multi-energy complementary DC microgrid isensured, the optimal operation of the system is ensured, and the influence caused by uncertainty is reduced; and the economy of the system is improved, and the working efficiency of the microgrid is improved.

The invention provides a micro-grid frequency control method with the lowest power generation cost, and belongs to the technical field of electric system operation and control. The method comprises the steps that a respective power generation cost function and a power generation cost micro increasing rate function are established; active power output limit constraints of distributed power supplies and energy constraints of energy storage devices are set; and the synchronization speed regulator of a power generator and the controller of an inverter are regulated to perform primary frequency modulation, secondary frequency modulation is performed after a time interval, system frequency is acquired and active power set value total regulating amount of each time of secondary frequency modulation is calculated, the active power set value total regulating amount is allocated for all the distributed power supplies or the energy storage devices participating in frequency modulation, the active power set value regulating amount and the original set values are summated to act as a new active power set value, then one time of frequency modulation is performed again and stable system frequency can be maintained. Utilization rate of renewable energy is enabled to be the maximum by the method, and consumption capacity of micro-grid renewable energy can be enhanced. Besides, realization is convenient and control is easy.

The invention discloses a capacity optimization configuration method for an urban rail transit photovoltaic energy storage system. The method comprises the steps: formulating a time-sharing operationstrategy according to the operation characteristics that the morning and evening peak load capacities of urban rail transit are large to determine operation states of photovoltaic, energy storage andpower grid at different time periods; researching a typical daily load curve and a photovoltaic output curve to carry out output adaptability analysis, and establishing an optimization model taking voltage fluctuation and economy as target functions by integrating photovoltaic and energy storage time-sharing operation characteristics and output adaptability analysis; converting multiple targets into a single-target optimization function in a weight assignment mode; carrying out solving through a particle swarm optimization algorithm to obtain optimal configuration of the energy storage capacity of the system. The method facilitates energy conservation and emission reduction of the urban rail system and local photovoltaic absorption, meanwhile, reduces network voltage fluctuation to improvethe traction side electric energy quality; and time-sharing operation is achieved, photovoltaic peak clipping and valley filling and load peak clipping are conducted through energy storage, and the economical efficiency of the system can be improved.

The invention provides an active power coordination control method and system of an optical storage power station. The method comprises the following steps: calculating the reference value of the active power of the optical storage power station according to the frequency of the optical storage power station, the frequency change rate and the active power control instruction of the automatic generation of the electric network; determining The active power coordination control mode of optical storage power station according to the frequency of the optical storage power station. Based on the active power reference value and the active power coordination control mode of the photovoltaic storage power station, obtaining the active power control commands of the photovoltaic unit and the energystorage unit, and carrying out the active power regulation of the photovoltaic storage power station according to the active power control commands of the photovoltaic unit and the energy storage unit. The technical proposal provided by the invention improves the power grid stability support strength by configuring an energy storage system, considering the use of an energy storage unit with maximum capacity, and carrying out centralized active power coordination control on the photovoltaic unit and the energy storage unit.

The invention discloses a reactive power optimization method for a photovoltaic reactive power partition pricing power distribution network, and the method comprises the steps: proposing a controllable photovoltaic reactive power partition pricing model, taking the minimum reactive comprehensive cost and the minimum deviation between active power loss and voltage as optimization targets, and constructing a power distribution network active-reactive power optimization model including response PV inverter reactive partition electricity price, static Var compensator (SVC) dynamic compensation, on-load voltage regulation equipment OLTC tap and capacitor bank switching; and processing nonlinear terms in the model by adopting mathematical means such as second-order cone programming and a large Mmethod, finally obtaining a mixed integer second-order cone programming model of active-reactive coordination optimization of the power distribution network containing controllable photovoltaic in the power market environment, and segmenting and solving the model by adopting a Benders decomposition method. According to the invention, the reactive capacity of a photovoltaic inverter and the reactive support capability of original discrete equipment of a distribution network are reasonably and coordinately utilized, the active power loss is reduced, and the user voltage quality is ensured.

The invention discloses an active distribution network planning method, comprising the steps of establishing a photovoltaic power generation output model and reducing an operation scene; Establishingthe optimization model of distribution company grid construction; Establishing the electricity price regulation model of distribution company; Generating a distribution network grid structure; obtaining the final active distribution network planning scheme by solving the generated model. The method of the invention can solve the problem of active distribution network frame expansion planning, canselect the optimal distribution network line construction scheme, reduce investment cost, improve distributed power dissipation capacity, reduce light discarding, effectively reduce network loss, andimprove the economy of distribution network.

The invention provides a hydrogen-containing energy storage microgrid optimization operation method based on deep reinforcement learning, and the method comprises the steps: building an electrolytic cell efficiency characteristic model through a linear interpolation method, combining the electrolytic cell efficiency characteristic model with a gas turbine model, an electrochemical energy storage model, a hydrogen storage tank model, a fuel cell model and the like, and taking the microgrid operation cost as a target, constructing an optimized operation model of the hydrogen-containing energy storage micro-grid. And finally, solving a sequence decision problem of micro-grid optimization operation by adopting a depth deterministic strategy gradient algorithm. According to the method, the efficiency characteristic of the electrolytic cell is considered, the hydrogen energy storage capacity can be fully utilized, the optimization problem is solved according to the deep reinforcement learning principle, the operation cost of the hydrogen-containing energy storage microgrid is reduced, and good generalization is achieved.

The invention discloses a garden integrated energy system day-ahead economic dispatching method considering wind and light uncertainty. Firstly, a two-stage adjustable robust optimization day-ahead dispatching model considering combined thermoelectric demand response is established by taking minimum basic scene operation cost as an objective function and combining related equipment and operation constraints of a garden system. due to the fact that the CCG method has better calculation efficiency in processing the robust optimization model, the CCG method is used for solving. And finally, a commercial solver Gurobi is applied to perform example simulation, a two-stage robust optimization scheduling strategy considering the combined thermoelectric demand response is obtained, it is verified that the method can better process uncertainty occurring in the system, meanwhile, the operation cost of a park can be reduced, and consumption of new energy is promoted.

The invention relates to the technical field of solar thermal power generation, and discloses a thermochemical heat storage system and method. The thermochemical heat storage system comprises a low-temperature fused salt storage tank, a heat absorber and a thermochemical fused salt storage tank which are connected in sequence and form a circulation loop. The thermochemical fused salt storage tank is provided with a thermochemical reactor, a fused salt circulation pipeline is arranged in the thermochemical reactor, the thermochemical reactor is filled with a thermochemical heat storage module, and the thermochemical heat storage module can be subjected to the dehydration reaction at the high temperature and absorb heat. The thermochemical heat storage system is further provided with an evaporator and a reaction gas pipeline communicating with the evaporator and the thermochemical reactor. Fused salt flowing out of the low-temperature fused salt storage tank is heated in the heat absorber, then enters the thermochemical fused salt storage tank, exchanges heat with the thermochemical heat storage module and is used for the dehydration reaction of the thermochemical heat storage module, heat is stored in the thermochemical heat storage module in the form of chemical energy, the heat storage efficiency is improved, abandoned light is reduced, and the utilization rate of solar energy resources is increased.

The invention relates to a photovoltaic grid-connected inverter control method, in particular to a photovoltaic grid-connected inverter control method for preventing voltage out-of-limit of a low-voltage distribution network by considering the influence of line impedance and operating environments on a photovoltaic inverter. Voltage out-of-limit of a grid connection point of the inverter can be prevented, and light discarding amount can be decreased as far as possible by the aid of the residual capacity of the inverter on the premise of ensuring the voltage quality of the grid connection point when voltage exceeds upper limit. Compared with a traditional method, when the voltage of the grid connection point exceeds the upper limit, an active and reactive power matching method is designed by reverse thinking on the premise of ensuring the voltage quality, and the light discarding amount is decreased by the residual capacity of the inverter.

The invention relates to a wind-light-fire extra-high-voltage direct-current delivery scheduling method. The method comprises the following steps: determining a first weight coefficient of peak regulation time periods of equivalent new energy and a receiving-end power grid load, a second weight coefficient of a source load disturbance degree and a third weight coefficient of direct-current delivery new energy according to a fuzzy analytic hierarchy process; determining the minimum thermal power capacity; taking the lowest peak regulation degree of a receiving-end power grid as a target, and constructing a target function according to the peak regulation time period, the source load disturbance degree and the weight coefficient thereof; and constructing constraint conditions according to the minimum thermal power capacity, the thermal power generating unit output data, the new energy output data, the direct current channel output data and the total preset electric quantity of the directcurrent delivery day transaction, and finally obtaining an optimal solution of optimal scheduling. By optimizing the extra-high-voltage direct-current delivery power, the comprehensive use capacity of new energy can be improved, the reverse peak regulation time period and the source load disturbance degree of the receiving-end power grid are reduced, the wind curtailment and light curtailment amount is reduced, the peak regulation capacity of extra-high-voltage direct-current delivery wind, light and fire is improved, and the peak regulation margin of the receiving-end power grid is improved.

The invention discloses a multiple flexible resource multi-state risk calculation method. A single flexible resource schedulable potential value evaluation analysis model is established for a power grid comprising an electric automobile, a heat pump, an air conditioner and a distributed power source, and schedulable potential values of the electric automobile, the heat pump, the air conditioner and the distributed power source can be respectively acquired through the single flexible resource schedulable potential value evaluation analysis model; a multiple flexible resource gathering schedulable potential value evaluation model is established, and a schedulable potential value after multiple flexible resource gathering is acquired by calculation, a multiple flexible resource multi-state risk calculation model is established, different probabilities of the multiple flexible resources in multiple states are acquired, and the multiple flexible resource multi-state risk information is acquired through calculation. The method is advantaged in that the flexible resource risk information of the power grid is acquired under the action of multiple uncertain factors, a risk level of the power grid under the action of uncertain factors is identified, bases are provided for bidirectional interaction between the flexible resources and the power grid, and scientific bases are provided for safe operation of an intelligent power grid.

The invention provides an electric power peak shaving method, a dispatching center and an electric power peak shaving system. The method comprises the following steps: obtaining next-day predicted load, next-day predicted new energy output and next-day high-energy load information capable of participating in dispatching; generating a day-ahead scheduling plan under a peak regulation mode in whichonly a conventional energy power plant unit participates according to the obtained next-day predicted load, the next-day predicted new energy output and the next-day high-energy load information capable of participating in dispatching; according to the day-ahead scheduling plan, judging whether a case of discarding new energy output exists the next day; if so, generating a day-ahead scheduling plan in which the conventional energy power plant unit and high-energy load jointly participate; and carrying out intra-day correction on output plans of a conventional energy power plant and a new energy power plant based on the day-ahead scheduling plan and the battery energy storage device. Capability of tracking load change can be enhanced; the adjustment flexibility can be enhanced; the limitation of high-energy load discrete adjustment on peak regulation can be reduced; and the system operation cost can be effectively reduced.

The invention discloses a network source coordination virtual synchronous machine control method of an optical storage integrated system. An optical storage integrated system based on the method is ofa two-stage structure. Photovoltaic and energy storage are connected to a power grid by adopting the same inverter, photovoltaic and energy storage batteries are respectively connected to a direct current bus of the inverter in parallel through respective DC / DC converters, and the two DC / DC converters and the inverter are controlled by adopting a network source coordination virtual synchronous machine control scheme. The inverter realizes inertia control and reactive droop control; a DC / DC converter of the battery realizes active droop control; when the voltage of the direct-current capacitoris not higher than the upper limit value Udcuplimit, the photovoltaic DC / DC converter realizes maximum power output control; and when the voltage of the direct current capacitor exceeds the upper limit value Udcuplimit, the photovoltaic DC / DC converter automatically enters light abandoning control. Inertia and droop support can be provided for the power grid, and the power output to the power grid can be automatically adjusted when the source side power fluctuates, so that light abandoning is reduced, and the requirement for the capacity of an energy storage battery is reduced.

The invention discloses an active distribution network planning method, which includes establishing a photovoltaic power generation output model and reducing operation scenarios; establishing an optimization model for grid construction of a power distribution company; establishing an electricity price adjustment model for a power distribution company; and generating a grid structure of a distribution network ; Solve the generated model to get the final active distribution network planning scheme. The method of the present invention can solve the problem of active distribution network grid expansion planning, can select the optimal distribution network line construction scheme, reduce investment costs, improve the consumption capacity of distributed power sources, reduce light abandonment, and effectively reduce network loss. Improve the economics of the distribution network.

According to the comprehensive energy system operation optimization device considering the user demand response characteristics under the carbon emission cost, based on a computer system, the carbon emission characteristics of a coal-fired power generation unit and a gas-fired boiler are analyzed, a comprehensive energy system carbon dioxide emission model and a stepped carbon emission cost model are constructed, and the user demand response characteristics of the coal-fired power generation unit and the gas-fired boiler are optimized. Calculating the carbon dioxide emission cost of energy consumption at each moment; on the basis, a user demand response model guided by double signals of the carbon emission cost and the energy purchase price is constructed, and an integrated energy system operation scheduling model is constructed with the purpose of minimizing the energy purchase cost and the carbon emission cost of the integrated energy system. The system can fully reflect the environment cost of operation of coal electricity, fuel gas and other equipment, more effectively encourages a user to use more energy in the clean energy output peak period, promotes the consumption of clean energy, reduces the wind and light abandoning conditions of the system, and reduces the degree of dependence on external energy at the same time.

The invention discloses a method of improving tracking photovoltaic power generation output characteristics based on hydrogen and battery hybrid energy storage. The method comprises the steps of: S100, formulating a next-day photovoltaic output plan according to the short-term photovoltaic prediction power, and correcting the next-day photovoltaic output plan according to the SOC of an energy storage battery at night of the day, S200, dividing scheduling modes into main scheduling and coordinated scheduling according to the SOC of the energy storage battery, selecting a different scheduling mode according to the state of charge interval of the SOC of the energy storage battery in real-time scheduling, and tracking the photovoltaic planned output, and S300, taking the photovoltaic output plan as the premise by next-day scheduling and controlling a hydrogen electrolysis cell and the energy storage battery for charge and discharge according to the scheduling mode. The method can perform flexible switching between a main scheduling scheme and a coordinated scheduling scheme, realizes accurate tracking of photovoltaic planned output, reduces fluctuation of grid-connected input power ofa photovoltaic power station, relieves frequency modulation pressure of a system unit, and improves the utilization rate of light energy.

The invention discloses a network source coordination virtual synchronous machine control method for an integrated optical storage system. The method is based on an integrated optical storage system with a two-stage structure, in which photovoltaic and energy storage use the same inverter to connect to the grid, and photovoltaic and The energy storage batteries are respectively connected in parallel to the DC bus of the inverter through their respective DC / DC converters, and the two DC / DC converters and the inverter are controlled by a virtual synchronous machine control scheme coordinated by the network source. The inverter realizes inertia control and reactive power droop control; the DC / DC converter of the battery realizes active power droop control; when the DC capacitor voltage is not higher than the upper limit value U dcuplimit When , the photovoltaic DC / DC converter realizes the maximum power output control; when the DC capacitor voltage exceeds the upper limit value U dcuplimit , the photovoltaic DC / DC converter automatically enters the control of light abandonment. The invention can not only provide inertial and droop support for the power grid, but also automatically adjust the power output to the power grid when the source side power fluctuates, thereby reducing light curtailment and reducing the demand for energy storage battery capacity.

The present invention proposes a micro-grid frequency control method that minimizes power generation costs, which belongs to the technical field of power system operation and control. The method includes: establishing respective power generation cost functions and power generation cost micro-increase rate functions; Power output limitation constraints, and energy constraints of energy storage devices: adjust the synchronous governor of the generator and the controller of the inverter, perform frequency modulation once, perform frequency modulation twice after time intervals, collect system frequency and calculate second frequency modulation each time The total amount of active power set value adjustments is distributed to all distributed power sources or energy storage devices participating in frequency regulation, and the active power set value adjustment amount and the original set value are summed as For the new active power setting value, perform frequency modulation again to maintain the system frequency stability. The method of the invention can maximize the utilization rate of renewable energy, can improve the absorbing capacity of the microgrid renewable energy, and is convenient to realize and simple to control.

The invention discloses an economic quantitative model for evaluating the improvement degree of a power distribution network project on the operation performance of a power distribution network. The model is used for evaluating the improvement degree of the operation performance of the power distribution network project based on project investment cost so as to help investors to analyze the cost performance of a target power distribution network project and the overall profitability of the society. The model comprises benefit increase brought by power grid performance improvement in four aspects of power distribution network power supply reliability improvement, power grid improvement, power supply capability enhancement and energy utilization enhancement caused by a power distribution network project, and project investment construction cost and later operation and maintenance cost. Economic quantitative analysis is performed on the power grid operation performance improvement broughtby the project based on the economic cost of the distribution network project so that the defect that the actual economic 'benefit' of the distribution network project is difficult to quantify due tothe fact that the performance improvement in the technical aspect of the project and the investment cost in the economic aspect are separately compared by the existing model can be compensated.

The invention provides a power peak shaving method, a dispatch center and a power peak shaving system. The method includes: obtaining the next day's predicted load, the next day's predicted new energy output, and the next day's high-energy load information that can participate in dispatching; according to the obtained The next day's forecasted load, the next day's forecasted new energy output, and the next day's high-energy load information that can participate in dispatching can generate a day-ahead dispatching plan under the peak-shaving mode in which only conventional energy power plant units participate; Whether there is a situation of discarding new energy output on a daily basis; if so, generate a day-ahead scheduling plan in which conventional energy power plant units and high energy loads participate together; based on the day-ahead scheduling plan and battery energy storage device output plan for conventional energy power plants and new energy power plants Intra-day correction can improve the ability to track load changes, increase the flexibility of adjustment, reduce the limitations of discrete adjustment of high-energy loads on peak regulation, and effectively reduce system operating costs.

The invention relates to an independent light-storage micro-grid topology for interconnected operation. The independent light-storage micro-grid topology is formed by connecting an energy storage system, a distributed photovoltaic power generation system and a load in parallel; and the distributed photovoltaic power generation system and the load are connected with an AC bus in parallel. The independent light-storage micro-grid topology is characterized in that the output end of an energy storage battery pack in the energy storage system is connected to one end of an isolating switch; the other end of the isolating switch is connected with the input end of a DC / DC converter; the output end of the DC / DC converter is connected with the DC side of a DC / AC energy storage converter, and the connecting line is taken as a DC bus of independent light-storage micro-grids; and the AC output of the DC / AC energy storage converter is taken as an AC bus of independent light-storage micro-grids. When the independent light-storage micro-grids work independently, the DC / DC converter controls the voltage of the DC bus of the independent light-storage micro-grids, and the DC / AC energy storage converter controls the voltage of the AC bus of the independent light-storage micro-grids; when the independent light-storage micro-grids are interconnected, the two independent light-storage micro-grids are connected in parallel by utilizing an interconnecting switch; and the DC / DC converter controls the voltages of the micro-grids by adopting an active voltage droop control method.

The invention discloses a photovoltaic power station bidding optimization method based on consideration of an energy storage demand response effect. The method comprises the steps of obtaining a comprehensive price incentive value of a system operator for a photovoltaic output curve based on preset evaluation index data; calculating the expected electricity selling income of the photovoltaic powerstation based on the photovoltaic output basic rate, the comprehensive price excitation value, the bidding curve and other data, and calculating the cost of purchasing the energy storage resources with the demand response effect; and establishing a photovoltaic power station expected revenue maximization objective function, iteratively searching an expected revenue maximum value through an intelligent optimization algorithm, and obtaining a bidding curve and an energy storage effect demand response resource purchase strategy when the expected revenue maximum value is obtained. The incentive price mechanism is added into the expected revenue maximization objective function of the photovoltaic power station, so that the photovoltaic power station can be forced to purchase the energy storageresource with the demand response effect, the energy storage resource with the demand response effect is reasonably utilized and purchased, and power generation of the photovoltaic power station is more economical.

The invention provides a gas turbine system comprising a gas compressor, a combustion chamber, a turbine and heating devices, wherein the gas compressor comprises a gas compressor cylinder and a gas compressor rotor and is used for enabling a working medium to flow between the gas compressor cylinder and the gas compressor rotor and be compressed; the combustion chamber is connected with a workingmedium output end of the gas compressor; the turbine comprises a turbine cylinder and a turbine rotor and is connected with a working medium output end of the combustion chamber; and the heating devices are arranged on the outer surfaces of the gas compressor cylinder and the turbine cylinder and used for keeping the average temperature of the gas compressor cylinder and the turbine cylinder notlower than 50% of the average temperature of the gas compressor cylinder and the turbine cylinder in the stable working state of the gas compressor cylinder and the turbine cylinder so as to increasethe starting speed of the gas turbine system. The invention further provides a coupled power system which utilizes the gas turbine system and the new energy power generation system. The new energy power generation system is provided with a new energy set, a control system, a load and a power grid.

The invention discloses a distribution network reactive power optimization method for photovoltaic reactive power partition pricing, which includes proposing a controllable photovoltaic reactive power partition pricing model, with the minimum reactive power comprehensive cost, the minimum active network loss and voltage deviation as the optimization target, Construct active-reactive power optimization model of distribution network including response to PV inverter reactive power partition price, static var compensator SVC dynamic compensation, on-load voltage regulation equipment OLTC tap and capacitor bank switching; adopt second-order cone programming The nonlinear terms in the model are processed by mathematical methods such as the large M method, and finally a mixed integer second-order cone programming model for the coordinated optimization of active and reactive power of the distribution network with controllable photovoltaics under the power market environment is obtained, and the Bund The Sri Lankan decomposition method solves the model segmentation. The invention rationally and coordinately utilizes the reactive power capacity of the photovoltaic inverter and the reactive power support capability of the original discrete equipment of the distribution network, reduces active power network loss, and ensures user voltage quality.

The invention belongs to the field of electric power automation, and discloses a bidding capacity determination method, system and device for a virtual power plant and a storage medium, and the method comprises the steps of obtaining the predicted power generation power and prediction error distribution of an uncontrollable power supply bidding day in the virtual power plant, and the maximum power generation power of a controllable power supply bidding day in the virtual power plant; according to the prediction error distribution of the uncontrollable power supply bidding day, obtaining a plurality of reserved capacities of the controllable power supply bidding day and deviation power of each reserved capacity; solving a preset generated power determination model to obtain the maximum profit of the virtual power plant participating in the power market and the generated power under the maximum profit under each reserved capacity; and obtaining the reserved capacity with the maximum profit of the virtual power plant participating in the electricity market in all the reserved capacities, and taking the generated power under the reserved capacity as the bidding capacity. The maximum participation of the actual power generation power of the uncontrollable power supply is realized, the wind curtailment amount and the light curtailment amount are greatly reduced, the consumption of clean energy is effectively promoted, and the overall power generation cost is reduced.

The invention relates to a photovoltaic grid-connected inverter control method, in particular to a photovoltaic grid-connected inverter control method for preventing voltage out-of-limit of a low-voltage distribution network by considering the influence of line impedance and operating environments on a photovoltaic inverter. Voltage out-of-limit of a grid connection point of the inverter can be prevented, and light discarding amount can be decreased as far as possible by the aid of the residual capacity of the inverter on the premise of ensuring the voltage quality of the grid connection point when voltage exceeds upper limit. Compared with a traditional method, when the voltage of the grid connection point exceeds the upper limit, an active and reactive power matching method is designed by reverse thinking on the premise of ensuring the voltage quality, and the light discarding amount is decreased by the residual capacity of the inverter.

The invention discloses a capacity optimization configuration method for urban rail transit photovoltaic energy storage systems. According to the operation characteristics of urban rail transit with large morning and evening peak loads, a time-sharing operation strategy is formulated to determine the operation of photovoltaics, energy storage and power grids in different periods status; study typical daily load curves and photovoltaic output curves for output adaptability analysis, integrate photovoltaic and energy storage time-sharing operation characteristics and output adaptability analysis, establish an optimization model with voltage fluctuation and economy as the objective function, and pass the weighted value The multi-objectives are converted into single-objective optimization functions by means of a single-objective optimization function, and the particle swarm optimization algorithm is used to solve the problem to obtain the optimal configuration of the energy storage capacity of the system. The invention is beneficial to promote the energy saving and emission reduction of the urban rail system and the consumption of local photovoltaics, and at the same time reduce the network voltage fluctuations to improve the power quality of the traction side, time-sharing operation and use of energy storage to cut peaks and valleys of photovoltaics and load peaks Improve system economy.