Flexible expansion planning of distribution system integrating multiple renewable energy sources: An approximate dynamic programming approach

Abstract Distribution system expansion planning is a long-term process with sequential multistage uncertain factors, so flexible planning methods are required to deal with the various potential investment risks. This paper presents an approximate dynamic programming based flexible distribution system expansion planning model, in which the long-term system load growth uncertainty and short-term power fluctuation are both considered. A flexible investment strategy based on Markov decision process is developed, in which the planning decisions are made sequentially with consideration of both the current costs and future variable costs under long-term load growth uncertainty. This problem is formulated as a large-scale multistage stochastic programming model, which is intractable due to the "curse of dimensionality". Thus, an approximate dynamic programming approach is used to decompose the original multistage optimization problem into sequential subproblems that can be easily solved. Case studies are carried out on a 24-node distribution system with three planning stages and a 54-node distribution system with five planning stages. Numerical results validate the feasibility and benefits of the proposed planning approach in obtaining a flexible expansion scheme, which can significantly reduce the investment risk and configure renewable energy equipment more reasonably.