1.Smart Grid Research Institute, Nanjing Institute of Technology, Nanjing 211167, China;2.College of Energy and Electrical Engineering, Hohai University, Nanjing 211100, China

The correlation between distributed renewable energy and load has a crucial impact on the distribution network planning. Firstly, based on the truncated R-Vine Copula model, the joint probability distribution model of multiple loads, distributed wind power and photovoltaic generation is established. Then, a two-stage stochastic expansion programming model of the distribution network and its corresponding chance-constrained programming model are proposed. In this two-stage model, the first stage is the investment decision model of the distribution network, in order to minimize the total cost of the distribution network to determine the appropriate expansion and upgrade scheme of feeders, substations and capacitors. The second stage is the operation model of the distribution network in multiple scenarios. The objective of this stage is to minimize the network loss and the maintenance cost of the distribution network, and the optimal operation scheme is determined with the safe operation of the distribution network as the boundary condition. The correlation and randomness of loads and distributed generators are introduced into the model by the scenario and its realized probability. Considering that the solution of the scenario-based stochastic optimization model and the chance-constrained models is difficult, a new bilinear Benders decomposition algorithm is proposed, which can effectively solve the stochastic optimization model containing a large number of second-order cone power flow equations. Finally, taking a 38-bus distribution network in Nantong of China as an example, the proposed model and algorithm are verified.