CAREER: Integration of Planning and Control for Nonlinear Systems

道恩蒂尔伯里，美国密歇根大学9876039职业生涯：非线性系统的规划和控制的集成非线性系统的运动控制的经典方法可以分为两个独立的步骤。 首先进行路径规划，然后找到一个跟踪控制律，保证系统收敛到期望的路径。本研究的目标是发展一个统一的框架，在其中解决规划和跟踪问题的非线性系统。 实现这一研究目标的策略是首先量化非线性系统轨迹的几个特征。 轨迹规划算法，然后将开发一个广泛的非线性系统。 最后，将检查反馈控制与规划轨迹的集成，并开发用于生成规划和跟踪控制律的算法。 这项研究的结果将对自主系统的不断发展产生广泛的影响，从而能够更好地应对不断变化的条件，例如感知到的障碍物或大干扰。 通过综合考虑规划和跟踪问题，可以生成系统“容易”遵循的路径：快速，健壮，稳定，并且需要最少的能量。该项目的教育计划包括开发一套基于Web的教程，向学生讲授系统建模概念的重要性和实用性。 通过使用功能强大的仿真软件，非线性系统模型可以很容易地处理，更现实的例子可以纳入系统建模课程。

Dawn Tilbury, University of Michigan9876039CAREER: Integration of Planning and Control for Nonlinear SystemsThe classical approach to motion control for nonlinear systems can be broken down into two independent steps. Path planning is done first, and then a tracking control law is found which guarantees that the system converges to the desired path. The goal of this research is to develop a unified framework in which to address the both the planning and the tracking problems for nonlinear systems. The strategy for accomplishing this research goal is to first quantify several characteristics of trajectories for nonlinear systems. Trajectory planning algorithms will then be developed for a broad class of nonlinear systems. Finally, the integration of feedback control with the planned trajectories will be examined, and algorithms will be developed for generating both the planning and tracking control laws. The results of this research will have a broad impact on the growing area of autonomous systems, allowing for better reactions to changing conditions such sensed obstacles or large disturbances. By considering the planning and tracking problems in an integrated fashion, paths can be generated that are "easy" for the system to follow: fast, robust, stable, and requiring minimum energy.The education plan for this project includes developing a set of web-based tutorials to teach students the importance and utility of system modeling concepts. By using powerful simulation software, nonlinear system models can be treated quite easily, and more realistic examples can be incorporated into system modeling courses.