Ship pilotage constitutes a critical element of port safety, where the reliability of pilots is paramount to safeguarding the operational integrity and safety of both ports and vessels. Understanding the effectiveness of pilots under different levels of reliability is crucial for assessing pilotage safety. Therefore, this study first improves the cognitive reliability and error analysis method (CREAM) using decision-makingand evaluation experiment method (DEMATEL) and improved CPC factor influence rules. The enhanced CREAM method is subsequently employed to quantify human reliability. Furthermore, this study models the piloting process using the classical PID control framework and establishes a quantitative relationship between human reliability and PID parameters. Human reliability is considered as a factor influencing the control system, thus constructing a pilotage control model based on human reliability. To validate the model, this study utilizes ship arrival data from Qingdao Port to simulate two typical pilotage scenarios: straight-line and complex-path arrivals. The simulation evaluates the effectiveness of pilotage control across varying levels of reliability, reflecting different pilot performance scenarios. The results indicate that higher reliability leads to faster response times and smaller route errors. Specifically, pilotage response time for low-reliability scenarios is four times longer than that for high-reliability scenarios, and route error is twice as large as that for high-reliability scenarios.