Suspension systems are critical to the ride comfort and handling stability of vehicles, but traditional passive suspensions fail to achieve optimal performance in the two aspects. Semi‐active suspension controller changes the suspension stiffness or damping to improve the ride comfort and handling stability, which has the potential advantages of low energy consumption and high reliability. In this paper, the model of an air spring with an auxiliary chamber and a variable throttle orifice is proposed and the semi‐active control strategy of air suspension is carried out. The parameters of the air spring with auxiliary chamber are optimized to improve the performance when the control system fails and to increase the basic performance for semi‐active control. Based on the linear matrix inequality approach to robust H_∞ control, the semi‐active control strategy of the air suspension is studied, in which parameter uncertainties and time delay are considered to improve control robustness. An inverse model of the air spring is then established to calculate the area of the variable throttle orifice, by which the desired control force is tracked accurately. Finally, the effectiveness of the semi‐active control system of the air suspension is verified by numerical simulation. Comparing with the passive suspension, not only under nominal parameters, but also with parameter uncertainties, time delay, and both parameter uncertainties and time delay, the semi‐active control system proposed in this paper has good control performance as well as is strong robustness to parameter uncertainties and time delay.

中文翻译：

---

具有参数不确定性和时滞的带有辅助腔室的空气悬架半主动控制

悬架系统对于车辆的乘坐舒适性和操纵稳定性至关重要，但是传统的被动悬架无法在两个方面实现最佳性能。半主动式悬架控制器可以改变悬架的刚度或阻尼，从而提高乘坐舒适性和操纵稳定性，这具有低能耗和高可靠性的潜在优势。本文提出了一种带有辅助腔和可变节流孔的空气弹簧模型，并实现了空气悬架的半主动控制策略。优化带有辅助腔室的空气弹簧的参数，以改善控制系统故障时的性能，并提高半主动控制的基本性能。基于线性矩阵不等式的鲁棒H方法_∞控制，研究了空气悬架的半主动控制策略，其中考虑了参数不确定性和时延以提高控制鲁棒性。然后建立空气弹簧的逆模型，以计算可变节流孔的面积，通过该面积精确地跟踪所需的控制力。最后，通过数值仿真验证了空气悬架半主动控制系统的有效性。与被动悬挂相比，不仅在标称参数下，而且在参数不确定性，时间延迟以及参数不确定性和时间延迟方面，本文提出的半主动控制系统具有良好的控制性能，并且具有较强的鲁棒性。参数不确定性和时间延迟。