Model-and-data hybrid driven method for power system Operational reliability evaluation with high penetration of renewable energy

Abstract

With the rising penetration of renewable energy, the fluctuation of operating conditions such as wind speed and light intensity affects the reliability of the power system. It leads to the repeated reliability evaluation of the power systems and increases the computational complexity. Based on the non-sequential Monte Carlo simulation (NMCS) and back propagation neural network (BPNN), a new model-and-data hybrid driven method for power system reliability evaluation is proposed. Firstly, NMCS is used to calculate the system reliability indices under different operating conditions to obtain the input data for neural network training. Then, the proposed Multiple BPNN (M-BPNN) is used to establish the highly nonlinear mapping relationship between operating conditions and system reliability indices, which can significantly reduce the calculation time of power system operational reliability evaluation with high penetration of renewable energy. The IEEE-RTS 79 system is used to verify the effectiveness and efficiency of the proposed method, and sensitivity analysis is performed.