In the first season of ratoon rice harvesting, it was necessary to maintain appropriate stubble height according to varietal characteristics to ensure successful regeneration in the second season. When harvesting ratoon rice with existing grain combine harvesters in China, frequent adjustment of the header height was often required due to uneven field terrain. Currently, this operation is primarily performed manually by the operator, which not only increased labor intensity but also constrained harvesting efficiency. To address this issue, this study designed an adaptive header height control system for a ratoon rice combine harvester. First, a structural model of the header copying wheel for height control was established. Second, the hydraulic system of a 4LZ-1.6Z grain combine harvester was modified and optimized, and a control system simulation model was developed using Simulink to compare and analyze the performance of classical PID and fuzzy PID control strategies. Then, performance verification was carried out using response accuracy and response speed as test indicators. The results showed that the mean absolute error for classical PID control was 17.3 mm, while that for fuzzy PID control was 10.5 mm. In terms of response speed, the average values for classical PID control during header lifting and lowering were 0.185 m/s and 0.213 m/s, respectively, whereas those for fuzzy PID control were 0.202 m/s and 0.219 m/s. Finally, a field test on uneven terrain was conducted based on the fuzzy PID control algorithm, demonstrating that the header height remained stable at approximately 500 mm, with a mean deviation of 6.81 mm and a stability coefficient of variation of 1.62%, indicating that the proposed control strategy exhibited good stability and tracking performance. This study provided an effective technical solution for header height control in the mechanical harvesting of ratoon rice, offering practical value for improving harvesting quality and operational efficiency.