【目的】水稻田间作业如旱直播和插秧对农业机械的路径控制精度提出了更高要求。为满足水稻生产中对作业一致性和作物均匀性的需求，本研究设计了一种用于轮式拖拉机的转向辅助控制系统，辅助农机驾驶员在水稻田间作业时实现精确的路径跟踪控制。【方法】首先，我们通过驾驶模拟器测试数据构建了驾驶员转向模型，并结合轮式拖拉机的运动学模型，开发了一个融合驾驶员在环的整车控制模型。然后，采用多胞线性参数时变系统来描述模型中的不确定性，包括时变的驾驶员模型参数和车速。基于此模型，设计了一种输出反馈鲁棒控制器，以确保系统在由多面体空间描述的不确定性下的鲁棒稳定性，此外，还采用了区域极点配置方法来提高系统的瞬态响应性能。最后，通过驾驶员在环测试对设计控制器的性能进行评估。【结果】结果表明，该方法能够有效提升轮式拖拉机的路径跟踪性能，同时减轻驾驶员的身体和心理负担，在单次车道变更操作和复杂蛇形道路两种不同行驶工况下，路径跟踪最大横向偏差较PID控制算法可以分别降低66.7%和55.6%，而驾驶员的物理负荷可以分别降低26.86%和37.6%，精神负荷可以分别降低4.2%和31.3%。【结论】所提出控制方法为提升水稻等作物田间作业质量提供了技术支撑。

【Objective】Field operations in rice cultivation, such as dry direct seeding and transplanting, require higher path tracking accuracy of agricultural machinery. To meet the requirements for operational consistency and uniform crop distribution in rice production, we designed an assisted steering control system for wheeled tractors, aiming to assist human operators in achieving precise path tracking during rice field operations.【Method】First, we constructed a driver steering model using data from driving simulator tests, and combined it with the kinematic model of wheeled tractors to develop an integrated vehicle control model that incorporates the driver-in-the-loop. Then, we used a polytopic linear parameter varying system to describe the uncertainties in the model, including time-varying driver model parameters and vehicle speed. Based on this model, an output feedback robust controller was designed to ensure robust stability of the system under uncertainties described by polyhedral sets. Additionally, a region pole placement method was employed to improve the transient response performance of the system. Finally, driver-in-the-loop tests were conducted to evaluate the performance of the designed controller.【Result】The results show that this method effectively enhances the path tracking performance of wheeled tractors while alleviating the physical and mental burdens on the operator. In single lane-change maneuvers and complex serpentine driving conditions, the maximum lateral tracking error is reduced by 66.7% and 55.6% compared with that of the PID control algorithm, respectively, while the physical workload of the driver was reduced by 26.86% and 37.6%, and the mental workload was reduced by 4.2% and 31.3%, respectively.【Conclusion】The proposed control strategy offers a technical foundation for enhancing the operational quality of rice and other crop fieldwork.