Research On Planning Of High Resilience Active Distribution Network

Increasingly frequent extreme weather disasters pose a great threat to the safe and stable operation of the power grid.It is imperative to build a strong power grid with strong resistance to extreme disasters and strong ability to restore power supply after disasters.The research on resilience of power grids aims to improve the power grid’s ability to cope with extreme weather disasters.This thesis conducts research on the planning of high resilience active distribution network,which has important theoretical and practical significance for the planning and construction of high resilience distribution network and improving the disaster resistance of distribution network.In this paper,the current research status in related fields are first reviewed.Then the cyberphysical system(CPS)architecture of active distribution network operation optimization is analyzed and the data-driven stochastic optimization decision-making algorithm for distribution network operation is studied.And then,the measures to improve the resilience of the active distribution network are defined as resilience reserves,the resilience reserves are classified and modeled,on this basis,a high resilience distribution network planning method based on the constraints of resilience reserves is proposed.Finally,the resilience reserve planning of the high resilience active distribution network based on deep reinforcement learning(DRL)is studied on the simulation example.The innovative results achieved in the article mainly include the following contents:1、An active distribution network resilience reserve model is established.According to the mechanism of resilience reserve in the process of distribution network failure recovery,energy support type,energy channel type and demand response type resilience reserve are classified and modeled.The resilience reserve comprehensively describes the synergistic effect of the measures at distribution network source-grid-load level on the improvement of resilience,and provides constraint conditions for the planning of high resilience active distribution network.2、A high resilience active distribution network planning method based on resilience reserve constraints is proposed.On the basis of obtaining high resilience planning goals and forming resilience reserve constraints,a double random scene with random failures and random operation is established to simulate the failure recovery process.Based on the stochastic optimization decision-making of markov decision process(MDP),a high resilience active distribution network planning model is established.The method overcomes the limitations of traditional disaster-resistant planning methods.3、The optimization decision result of resilience reserve is calculated using the DRL algorithm based on dueling deep Q network(DDQN),and the resilience reserve planning scheme of the high resilience active distribution network is obtained through the analysis of its impact on the improvement of resilience.A typical example of IEEE33 distribution network is used as a simulation.The simulation results verify the effectiveness of the proposed planning method in improving resilience.