Prediction Techniques for Accelerating the Convergence of Alternating Solution Method during the Power System Dynamic Simulation Using SDC Integration Algorithm

The electric power of generator P_e and bus voltage of network V have a strong nonlinear coupling relationship with other variables, which results in bad effects on the convergence procedure for the alternating solution method during the power system dynamic simulation using the SDC integration algorithm. Deep research was conducted into this problem. The convergence procedure for P_e and V was illustrated during the simulation, and prediction techniques by applying the Lagrange interpolation to construct a better iterative initial value with smaller error for them was proposed. Based on the SDC simulation method, the proposed method was tested on an IEEE standard 3-generator 9-bus system on the platform of MATLAB, and results show that the prediction techniques performs quite well for the acceleration of convergence procedure P_e for V and and simulation efficiency of SDC integration algorithm.