Distributed Generation Integration: Economic Planning Approach for Active Distribution Networks

In response to the issue that traditional distributed generation planning methods struggle to meet the zonal control requirements of future distribution networks with high proportions of distributed generation during the operation phase, this study first, a method for predicting photovoltaic power output based on the convolutional neural network and long short-term memory network is proposed to predict photovoltaic (PV) output. Then, Based on the prediction results, a two-layer collaborative location selection and capacity planning model for distributed photovoltaics and energy storage is constructed. Among them, the upper-layer model sets the annual comprehensive cost minimization as the objective function, and the decision variables involved include the installation number of distributed power generation and energy storage equipment; the lower-layer model takes the annual comprehensive operating cost and the lowest value of voltage offset as the objective function, and its decision variables cover the output of distributed power generation and energy storage, the power purchase from the main network, and the demand side response. Furthermore, considering the characteristics of this model, an improved whale optimization algorithm is adopted for solution. Finally, validation is conducted on the IEEE 33-bus system to verify the feasibility of the proposed model and the effectiveness of the solution method.