During helicopter operations for target tracking tasks， pilots perceive the current flight state through visual and motion cues， thereby generating corresponding control behaviors. To study pilot control behavior under different motion feedback modes， this paper conducts pilot-in-the-loop simulation experiments under two conditions： Without motion feedback and with roll angle motion feedback. A hybrid optimization algorithm combining genetic algorithms and the Gauss-Newton method is proposed for pilot model parameter estimation. Furthermore， the effects of different motion feedback modes on pilot control behavior and pilot model parameters are explored using variance analysis. The research results indicate that the pilot model parameters obtained using the hybrid optimization algorithm can effectively represent the pilot’s actual control behavior and dynamic characteristics. Under different motion feedback modes， pilots adjust their control strategies to improve performance in target tracking tasks and regulate their control activities.