Optimum planning of large distributed resources in a mesh connected system based on artificial neural networks

This paper proposes a new approach to optimally determine the appropriate size and location of a dispatchable large Distributed Resource (DR) in a large mesh connected system. Inserting a DR in an already existing distribution system is an important issue at present, specifically under the deregulated electricity market. Determining the optimal siting and/or sizing of the DR is the key factor in determining the penetration level of the renewable energy sources. Various parameters; like losses, voltage profile, etc were investigated in the previous work. Beside the losses and the voltage profile, the proposed approach introduces another important parameter related to DR installation, which is the short circuit level representing the capacity of the already existing network protective devices. The proposed algorithm uses Artificial Neural Networks (ANN) to determine the appropriate weighting factors of each parameter included in the optimization problem. The selected parameters, i.e., the voltage level, the total system losses and the short circuit level are weighted in order to choose the optimal DR allocation and its corresponding sizing. The proposed technique has been tested on the IEEE 24-bus mesh connected test system. This test system is a large heavily loaded interconnected system. The main advantages of the proposed optimization technique are its simplicity, and applicability to other systems, i.e. radial and interconnected. Simulation results using the MATLAB simulation package are presented to validate the effectiveness of the proposed approach.