Reactive Voltage Optimization Strategy Based on Improved Depth Deterministic Strategy Gradient Algorithm
Reactive voltage optimization is a necessary means to regulate the voltage and ensure the safe,stable and high quality operation of power system.A voltage control strategy based on the improved depth deterministic strategy gradient algorithm is proposed,aiming at the prominent voltage control contradictions and the difficulty of reactive power optimization in power systems.Firstly,the objective function of minimum network loss of power system is established,Markov decision process is used to model the reactive power optimization problem of power system,and Ornstein-Uhlenbeck(OU)process is introduced to generate autocorrelation noise,so that the agent can ensure the exploration in one direction and improve the learning efficiency.Secondly,the priority experience playback pool of Sumtree structure is used to improve the utilization of training samples,and importance sampling is used to optimize the convergence results.Finally,through the example of IEEE30-node standard system,it is verified that the method proposed in this paper can reduce the average network loss by 19.64%compared with the previous system,which can effectively reduce the active power loss of power grid and meet the needs of the development of power system.
NETL
NSTL
万方数据
