"Source-Network-Load-Vehicle-Storage" Collaborative Operation Optimization For Microgrid In Industrial Park

With the development of industrial modernization,more and more industrial parks have become the main body of energy production and consumption,and the mode of energy production and use in industrial parks has an important impact on the development of urban energy system.The direct connection of the distributed power supply,electric vehicles and other equipment in the industrial parks to the microgrids has a significant impact on their operation.The essence of the "source-network-load-vehicle-storage" collaborative operation optimization for microgrids in the industrial parks is to reasonably coordinate the power of the power generation and consumption of various units,power supply and power consumption equipment,make full use of the optimized capacity the equipment connected to the microgrids in a coordinated manner,improve the safe and stable operation level of the microgrids,and promote the economic benefits of both suppliers and buyers.It can not only reduce the cost of operation and maintenance of the microgrids and the network loss,but also reduce the cost of electricity for households,improve the voltage quality of the microgrids,and effectively ensure the economic and safe and stable operation of the microgrids.This paper takes the "source-network-load-vehicle-storage" collaborative operation optimization for microgrids in industrial parks as the research object,and summarizes as follows:(1)This paper studies the "source-network-load" collaborative operation optimization for microgrids in the industrial park.According to the characteristics of the industrial park load,considering the controllability of the air-conditioning load in the industrial park,based on the established controllable load model,and the multi-objective Q-learning algorithm and the non-dominated sorting genetic algorithm are used to solve the collaborative operation optimization model with the objects of minimizing the cost of operation and maintenance and the cost of carbon trading in microgrids.The "source-network-load" collaborative operation optimization model is used to coordinate the power output among power source,connecting power grid and controllable load,and the controllable load capacity in the industrial park is fully utilized to optimize the controllable load output in different time periods and adjust the load level of the microgrids under the premise of always meeting the comfort temperature requirements of users,and the model is suitable for the optimization scene under different weather conditions.Compared with non-dominated sorting genetic algorithm,multi-objective Q learning algorithm can make full use of the experience learned from the environment,optimize the action selection,enhance the directivity of the optimization process,improve the optimization ability of the optimization process,and increase the diversity of non-dominated solutions.Through sampling analysis,the multi-objective Q-learning algorithm can get more satisfactory results than the non-dominated sorting genetic algorithm in solving the "source-network-load" collaborative operation optimization model.and the multi-objective Q-learning algorithm can obtain more satisfactory solutions.(2)This paper studies the "source-network-load-vehicle" collaborative operation optimization for microgrids in industrial parks.Based on the load level of the microgrids and the capacity of charging and discharging of electric vehicles,a mathematical model of dynamic tariff of electric vehicles is established.The uncertainty of the microgrids photovoltaic and wind power generation in the industrial park is considered,and the stochastic problem of wind and photovoltaic power generation is solved by using the stochastic chance constrained programming method.The multi-objective collaborative operation optimization model of "source-network-load-vehicle" of the microgrids is established with the objectives of minimizing the load fluctuation rate of the microgrids and minimizing the charging and discharging cost of electric vehicles users.Compared with time-of-use tariff model,the dynamic tariff model of the EV is more pertinent when guiding EVs,which can strengthen the discharge behavior of EVs in peak load and the charging behavior of EVs in valley load,and reduce the discharge of EVs in peak and the charge of EVs in valley.EV dynamic pricing mode strengthens the guidance and management of EVs,enhances the stability of microgrids operation,and increases the income level of EV users participating in microgrids collaborative operation optimization.The dynamic tariff strategy is applicable to the guidance and optimization of EVs in different scenarios,and the load level and usage of EVs in the microgrids can adaptively determine the charging and discharging tariff of EVs,and the dynamic tariff strategy is convenient and effective.(3)This paper studies the "source-network-load-vehicle-storage" collaborative operation optimization for microgrids in industrial parks.Adding cooling,heat,and power storage systems to the microgrids collaborative operation optimization model,and considering the randomicity and fuzziness of photovoltaic power and wind power,the stochastic fuzziness of distributed power is treated by stochastic fuzzy chance constrained programming.The multi-objective collaborative operation optimization model of the microgrids is established with the objectives of minimizing the maximum node voltage deviation and minimizing the active losses in the microgris.The proposed collaborative operation optimization model can easily and reasonably coordinate the "source-network-load-vehicle-storage" power distribution of microgrids under different scales,make full use of and coordinate the capacity of the cooling-heat-electric energy storage system,reduce the waste of reheat power in industrial parks during the low cooling load and low heat load period,improve the energy supply capacity during the peak load period,and improve the flexibility and economy of microgrids and the power quality of microgrids.