Optimization methodology for control parameter of PI based on chaos prediction of electric powertrain

The electric powertrain is a nonlinear dynamic system, when electric vehicles (EVs) drive under torque control mode, unexpected oscillation of direct-axis current id and quadrature-axis currents iq of PMSM may occur in case of an unreasonable control parameter of PI regulator is set. Thus it influences effective and stable output torque of electric powertrain. The objective of this paper is to optimize the control parameter of PI regulator to improve the precision and stability of output torque of electric powertrain under torque control mode. To achieve the goal, nonlinear dynamic model of electric powertrain under torque control mode was established. Then we predicted the instability domain of control parameter of PI to instruct the optimization. Finally experimental platform was set up to operate the optimization. The optimization results were given to improve the precision and stability of output torque of electric powertrain. The results show that the output torque of PMSM continues to oscillate under EV start-up working condition and limited local oscillation occurs under low speed cruise condition. The electric powertrain cannot perform precise and stable torque control. The control parameter of PI was optimized from kp = 0.1 to kp = 0.18. The electric powertrain perform more stable torque control with more effective output torque after optimization.