| At present,the global environmental pollution,climate variation,and the increasing shortage of traditional fossil fuels have motivated the rapid development of distributed generation technology,especially the renewable energy generation technology.On the user side,electric vehicles have attracted the world’s attention to clean vehicles.In addition,due to the development of the advanced measurement technology as well as information and communication technology,power users are not only electricity consumers,but also participators in power grid operation.They analyze electricity information and adjust their own electricity usage behaviors accordingly.In the meantime,the development of Internet of things technology also makes smart home controllable.The smart home is represented by air conditioning,water heater,washing machine and lighting.When abundant distributed energy is connected to the distribution network,it will gradually evolve from the traditional passive network to an active system with complex features,bringing new challenges to the operation and dispatching of distribution network.Hence,this paper mainly examines the day-ahead optimal dispatching model and method of distribution network adapted to the new situation.Through coordinatively optimizing multiple types of distributed energy,the goals of elevating consumption of renewable energy,enhancing energy efficiency and improving load curve can be realized.The main contents of this paper are described as follows:The multi-source characteristics of active distribution network are studied.First,the output characteristics of wind power,photovoltaic power,gas turbine,fuel cells,and distributed energy storage are analyzed and the output models of distributed generation and the energy storage system for active distribution network are established.For flexible and controllable load demand side resources,we focus primarily on investigating the modeling method of electric vehicles and thermostatically controlled loads,establishing the random charging model of electric vehicles,examining the modeling method and control mode of thermostatically controlled loads,and building the demand response model of air conditioning loads.Finally,to examine the load regulation performance of flexible and controllable loads,this paper takes air conditioning loads as an example.Joint scheduling optimization method of the power generation side and demand side is putted forward and the power system scheduling model considering large scale wind power and the controllable load is built.The model is solved by CPLEX and the peak load shifting potential of air-conditioning load is analyzed.Building upon the above research on the multi-source characteristics of active distribution network,combined with the technology of virtual power plant,a virtual power plant model of large-scale air-conditioning load group is established.Accordingly,"generation-storage-load" coordinative optimal dispatching method of active distribution network is studied.From the perspective of improving the consumption of renewable energy,an optimal dispatching model of distribution networks is built,aiming to minimize renewable energy curtailment,the operational costs,and load curve variance.Based on Pareto optimal principle and fuzzy decision method,a multi-objective particle swarm optimization algorithm is proposed to solve the proposed model.Finally.the proposed model is simulated on the Matlab platform by using the multi-objective particle swarm algorithm.Numerical simulation shows the feasibility and the rationality of the proposed optimal dispatching method of active distribution network.Regarding to the rapidly increased demand for electricity as well as the widening difference between load peak and valley for urban power grid,we further explore the schedulable potential of user side resources based on above research on "generation-storage-load" coordinative optimal dispatching of active distribution network.In addition,a research method of active distribution optimization dispatching based on load equilibrium entropy is proposed,aiming to reduce the power supply during peak period,improve load distribution balance and enhance energy efficiency.First of all.the time-of-use electricity price is implemented for different electricity users in the distribution network.Then,by using the information entropy theory,a new load equilibrium index is proposed to reflect the overall equilibrium of the load curve and the allocation optimization of electric energy.Based on this entropy,we built an optimal dispatching model to minimize the operational cost and maximize the load curve equilibrium.The simulation results illustrate that the proposed dispatching model can balance the demands between economy and load equilibrium,and achieve cost reductions of system operations,peak load curtailment,and efficiency improvements.At the same time,the correctness and superiority of the proposed load equilibrium entropy index are analyzed.Finally,the stochastic optimization dispatching method based on stochastic power flow modeling is studied.For the stochastic optimization scheduling problems brought by the high penetration of wind power and the connection of electric vehicles to active distribution network,the probabilistic models of wind power output,electric vehicle charging behavior and load power are established.Based on the existing stochastic power flow calculation method of distribution network,we adopt a stochastic power flow method based on unscented transformation to deal with the correlation of random variables.This method converts the uncertainty power flow calculation of abundant scenarios into the determination power flow calculation of slight Sigma sample points.On this basis,this paper proposes a novel multi-objective stochastic optimization model of active distribution networks that aims to minimize the operational costs,the power losses,and load variance.Finally,the proposed model is solved by using unscented transformation method and the particle swarm optimization algorithm.The simulation results show the correctness and rationality of the proposed method. |
PDF Full Text Request