An Optimization Model Based on Gradient Projection Method to Minimize Generation Cost in a Given Power Grid

In systems of power grid, controlling the supply from generators to match the demand is crucial. An Economic dispatch problem then aims to optimize the cost of power generation while producing enough power to supply into power grid. Non-Linear Programming (NLP) is utilized for optimizing such problems under constraints. However, current algorithms with high time complexity and less accuracy can result in unnecessary expenses and energy waste. This paper proposed an optimization model algorithm utilizing the Gradient Projection Method, which can be used to calculate real-time optimized results for each generator. The experimental results demonstrated that the proposed algorithm achieved superior performance in distributing precise amounts of power generation to various generators while maintaining a relatively short cost of time. The same algorithm can also provide results adapting to different power grid setup with high precision.