This paper compares various angle-tracking algorithms to balance the performance and noise for a steering-by-wire (SBW) system. Direct and quiet steering experiences can improve drivers’ acceptance of the SBW system. Linear quadratic regulator (LQR) control, robust control, and conventional cascade proportional–integral (PI) control have been developed and compared both theoretically and experimentally. To avoid the risky and time-consuming parameter-tuning process, a high-fidelity steering resistance model, which comprises a linear two-degree-of-freedom vehicle model and a dynamic LuGre friction model is established. Step and sine wave tests are simulated in a Matlab/Simulink environment to determine the reasonable parameter region for various methods. Then, the three types of algorithms are implemented on a prototype SBW vehicle and compared under the same scenarios. Finally, the simulated and experimental results are illustrated in detail. According to the indicators of control bandwidths, steady-state errors, cockpit sounds, and current waveforms, it is clear that LQR and robust control can achieve faster response and more acceptable noise, with uncertain and relatively larger tracking errors. Cascade PI control, in comparison, can realize smaller steady-state errors and gentler current waveforms, with slight noise and slower response.

本文比较了各种角度跟踪算法，以平衡线控转向 (SBW) 系统的性能和噪声。直接、安静的转向体验可以提高驾驶员对 SBW 系统的接受度。线性二次调节器 (LQR) 控制、鲁棒控制和传统级联比例积分 (PI) 控制已得到开发，并在理论上和实验上进行了比较。为了避免风险和耗时的参数调整过程，建立了高保真转向阻力模型，该模型由线性二自由度车辆模型和动态LuGre摩擦模型组成。在Matlab/Simulink环境中模拟阶跃波和正弦波测试，以确定各种方法的合理参数范围。然后，这三种算法在原型SBW车辆上实现，并在相同场景下进行比较。最后对模拟和实验结果进行了详细说明。根据控制带宽、稳态误差、座舱声音和电流波形等指标，可以看出LQR和鲁棒控制可以实现更快的响应和更可接受的噪声，但跟踪误差不确定且相对较大。相比之下，级联PI控制可以实现更小的稳态误差和更平缓的电流波形，噪声较小，响应速度较慢。