基于多移动机器人的持续监测控制研究

In many practical applications such as reconnaissance, search and rescue, and nuclear pollution monitoring, the monitoring objects are often located in dangerous and harsh environments. To accomplish the monitoring tasks without any casualties, multiple mobile robots systems are often used instead of humans. In current study, only finite points of interest in the mission domain are required to be monitored and it is assumed that the accumulation function defined on the points of interest grows or decreases linearly in time. In this project, a nonlinear accumulation function model is proposed based on our previous studies. The persistent monitoring problem is then addressed via a trajectory planning approach, where the robots' path planning is decoupled from speed control. The goal is to prevent the accumulation function from growing unbounded at any point of interest and persistent monitoring the whole mission domain simultaneously. Moreover, optimal control for multi-robot persistent monitoring is proposed to maximize the persistent monitoring frequency of the mission domain. This project is dedicated to further improving and enriching the control theory of persistent monitoring using multiple mobile robots, and providing new ideas and theoretical support for applications and spreading of multiple mobile robot systems.

在军事侦察、灾难搜救以及核污染监测等很多应用领域中，被监测对象往往处于危险与恶劣的环境中，采用多移动机器人系统代替人类执行持续监测任务可以避免人员伤亡，确保监测任务顺利进行。现有的基于兴趣点累积函数的多机器人持续监测研究均建立在兴趣点的累积函数随时间线性变化的前提之上，并且只考虑持续监测目标区域内有限个数的兴趣点。本课题在项目组前期研究基础上，提出随时间非线性变化的兴趣点累积函数模型，针对整个目标区域以及目标区域内的多个兴趣点，采用运动路径规划与速度控制器设计相分离的轨迹规划方法，确保所有兴趣点的非线性累积函数有界和对整个目标区域的持续性监测。在此基础上，进一步研究在持续监测频率性能指标下的多移动机器人持续监测优化控制问题。预期研究结果将丰富和完善基于多移动机器人的持续监测控制理论，为多移动机器人系统的实际应用和推广提供新的思路和理论依据。

本项目首先建立了更符合实际应用的随时间非线性变化的兴趣点累积函数模型。针对整个目标区域以及目标区域内的多个兴趣点，通过采用运动路径规划与速度控制器设计相分离的轨迹规划方法，确保所有兴趣点的非线性累积函数有界和对整个目标区域的持续性监测。在此基础上，进一步研究了在持续监测频率性能指标下的多移动机器人持续监测优化控制问题。对任务区域进行持续监测的一种有效的方法就是将移动机器人合理部署在任务区域使得给定的监测性能指标达到最优或次优。因此，项目组在完成本项目研究目标的同时，还深入研究了异构多移动机器人圆环覆盖控制问题，即如何设计移动机器人的分布式覆盖控制器最小化整个移动机器人网络到达给定圆环上任意一点的最长时间。本项目的研究结果将丰富和完善基于多移动机器人的持续监测控制理论，为多移动机器人系统的实际应用和推广提供新的思路和理论依据。