A Hybrid Neurodynamic Algorithm to Multi-objective Operation Management in Microgrid

Abstract

In this paper, we consider a microgrid framework consisting of four power generation units, such as gas turbine, fuel cell, diesel generator and photovoltaic power generation. We focus on the minimum power generation cost under the lowest environmental pollution, combining with particle swarm optimization (PSO) and projection neural network. In this framework, we consider the two objectives simultaneously, both economic cost and pollution emission. The projection neural network is used to find the local optimal value, and then the PSO algorithm is used to update the weight to increase the solution diversify and seek global optimization. The convergence and stability of the projection neural network algorithm are reflected in the simulation.