Aiming at the nonlinear and time-varying characteristics of plastic laser welding temperature control system, and the problems of low accuracy and large overshoot when using the traditional proportional-integral-derivative (PID) controller, an improved snake optimization (ISO) is proposed to optimize the PID parameters. Firstly, the elite reverse learning strategy is introduced into the snake optimization (SO) to generate the reverse solution, and new individuals are generated through comparison to improve the convergence speed of the algorithm; Simultaneously introducing the water wave dynamic adaptive factor to enhance the algorithm’s iterative optimization ability. By iterating the algorithm, a set of suitable PID parameters are obtained to control the temperature control system. The simulation results show that compared to traditional PID controllers, ISO-PID reduces the adjustment time by approximately 0.42 seconds and does not cause overshoot.