含执行机构汽车悬架系统一体化建模及自适应主动振动控制

Most of existing modeling work of active suspension systems for vehiculars did not consider the actuator dynamics and the embedded nonsmooth dynamics (e.g., friction, saturation and time-delay), and the associated active suspension control methods cannot address various constraints and suspension performance requirements simultaneiously. In order to address the above mentioned issues and promote the practical R&D of active suspension techniques, this project will study the modeling and parameter estiamtion of systems with nonsmooth dynamics, active suspension control designs and constrained multi-objective optimal control for vehicle suspension systems. The major work packages include: 1) Integrated modeling of active suspension systems with actuator dynamics and nonsmooth dynamics. We will reformulate the nonsmooth dynamics by investigating discontinuous piecewise parametric modeling (DPPM) technique, and then investigate novel adaptive laws with guaranteed fast convergence; 2) Adaptive active suspension control design with actuator dynanics. We will introduce simple, yet robust, unknown dynamics estimator to address the lumped unknown dynamics, and develop system cooridiante trannsform to investigate simplified control designs and implementation, which can regulate the motion of vehicular systems under road variations; 3) Constrained multi-objective adaptive optimal control designs. By proposing composite optimal performance function including constraints and multi-objective suspension performance requirments, we will formulate a new nonlinear optimal control to address active suspension, and then solve the derived optimazation equations online by using adaptive control techniques. This new framework will allow to address the constraints and several conflicting suspension requirments simultaneously. ..The objective of this project is to propose new modeling methods of vehicular active suspension systems with actuators and onsmooth dynamics, study adaptive active suspendion control designs and also exploit the use of adptive technique in solving the constrained multi-objective optimal control for suspension systems. Thus, it is paving a way for further theoretical developments of adaptive modleing and control of active suspension systems, and also for promoting the application of active suspension technology in practical vehicles.

面向汽车智能化及悬架系统高性能需求，针对目前主动悬架系统建模未完全考虑执行器及非光滑动态，控制器难同时满足多约束和多性能指标优化，造成在实际汽车应用困难的现状，研究含执行器非光滑动态主动悬架系统一体化建模、主动振动控制及多约束、多目标自适应优化控制，从建模、控制和优化角度探索改进主动悬架系统性能的方法。研究包括：1）含执行机构悬架系统一体化建模。引入分段参数化表达非光滑动态，构建一体化模型，研究快收敛自适应参数估计理论；2）含执行器主动振动控制设计。研究未知动态估计、系统等价转换等策略，实现路况干扰抑制；3）含约束多目标最优控制及实现。构建含约束和多目标优化综合性能指标，构建悬架最优控制问题并研究最优方程在线求解。研究目标为：提出含非光滑动态非线性系统一体化建模理论，建立自适应主动振动控制方法，提出含约束和多目标系统最优控制及求解策略，推进系统建模和控制相关理论研究和主动悬架应用进展。

本项目针对含执行机构的主动悬架系统精确建模和高性能控制的需求，从建模、控制和优化三个角度出发探索改进主动悬架系统综合性能的方法。主要的研究内容及研究成果为：1）针对忽视执行机构中非光滑动态而造成模型不精确的问题，引入分段连续参数化方法对非光滑动态进行参数化表征，建立了不依赖于使用中间变量和非光滑动态变量的主动悬架机构一体化模型，并构建了基于参数估计误差的自适应参数估计框架对系统未知参数进行估计，有效提升了主动悬架系统的建模精度。2）提出了一种结构简单、仅利用系统可量测输入、输出数据的未知动态估计器，实现了系统未知动态快速估计，并将估计得的未知动态估计前馈到控制器中以补偿非线性干扰，有效提升了主动悬架系统振动抑制性能。此外，还提出了预设瞬态收敛性能自适应控制系统的设计及分析方法，该控制方法不依赖任何函数逼近器，并可以有效适应系统未建模动态。将预设性能函数和误差转化纳入到控制器，实现瞬态和稳态位移严格收敛到预定范围内。3）针对车辆驾驶舒适性和减振装置安全性等要求，通过构建同时涵盖多个约束条件和性能指标的综合优化函数，并将主动悬架系统控制转化为最优控制问题，通过最优性原理推导出最优控制律，实现最优自适应控制器设计，改善悬架综合性能。4）上述理论成果在实验室搭建的全尺寸汽车主动悬架系统开展了应用研究和实验测试，验证了所提建模及控制方法的有效性和工程实用性。. 该项目研究成果系统性地建立了车辆主动悬架系统多约束和多目标最优控制器设计和分析方法，从一体化建模、主动振动控制和多性能指标优化控制三个方面开展基础理论和实验研究，并解决了主动悬架系统精确建模及高性能控制器设计等难题。本项目已有结果以及后续研究成果可作为非线性控制领域成果的有效补充，并对提升车辆主动悬架系统舒适性、可靠性和安全性，促进智能控制技术在汽车行业的应用，具有重要的现实意义和实用价值。