Abstract: With the development of renewable energy, the demand for external reactive voltage support in power systems has been steadily increasing. This paper establishes the mathematical model of the phasor excitation system, analyzing the influence of the time constants Ta, Tb, Tc of the PID controller and the gain Ke of the Automatic Voltage Regulator (AVR) on the voltage and reactive power response at the grid connection point. In response to the traditional Particle Swarm Optimization (PSO) algorithm"s tendency to prematurely converge and its slow convergence rate, a Logistic-chaos-initialized PSO is proposed. The controller parameters of the excitation system are optimized using the time-weighted absolute error integral (ITEA) as the fitness function. Simulation results show that the optimized phasor excitation system recovers voltage and reactive power more quickly, with smaller overshoot and reduced fluctuations after a three-phase short-circuit fault and subsequent disconnection, thus enhancing system stability.