Integrated Sensor System and its Application for the Navigation Control of Mobile Robots

本研究的目的是将不同工作原理的传感器组成一个集成的传感器系统，并将其应用于移动的机器人的导航控制中。本研究分别采用光学距离传感器和接触传感器来检测机器人与障碍物的距离和碰撞。在第一年的研究期间，所获得的结果是：（1）对于移动的机器人的导航控制，根据运行模式的级别划分任务并将它们应用于分步方法的控制是有用的。(2)如果传感器输出以模式获得，则直接使用该模式来控制机器人以避免碰撞是合适的。(3)相反，对传感器数据的基于规则的处理应用于机器人的逻辑导航，将模式数据和逻辑数据（符号数据）的处理结合起来，可以构成具有鲁棒性的导航控制系统。得到的结果是：（1）在实际的导航实验中，发现通过分级结构划分任务来控制机器人对于导航的鲁棒控制是有用的。(2)采用基于等价原则的路径规划方法，在通道型环境下进行了导航控制实验。机器人通过交叉路口时的距离传感器数据和机器人与墙壁碰撞历史的运行记录，检测到的有关运行情况的信息可以与预先在路线图上指示的信息相结合。(3)上述结果使得有可能通过整合距离传感器数据和触摸传感器数据来实现移动的机器人控制。

The aim of this research is to construct the integrated sensor system with sensors having different work principles and to apply the system to the navigation control of mobile robots.In this research optical range sensor and touch sensor are used for detecting the distance to the obstacle and the collision between robot and obstacles, respectively.The results obtained in the first year of research period are ;(1) For the navigation control of mobile robot it is useful to divide the tasks depending on the level of running mode and applying them for the control the step by step method.(2) If the sensor output is obtained in pattern it is suitable to use this pattern directly to control robots to avoid collision.(3) On the contrary the rule based processing of the sensor data should be used for the logical navigation of robot and by combining the processing of the pattern data and the logical data (symbolical data) the navigation control system with robustness can be constructed.Based on these results mentioned above navigation experiments have been done with the four-wheeled experimental robot manufactured in this project. The results obtained are ;(1) In the actual navigation experiment it is found useful to control the robot by dividing the tasks in hierarchical structure for robust control of navigation.(2) The navigation control experiments have been made in the passage type circumstance using pass planning based on cost-equal principle. The range sensor data obtained when robot is passing bythe crossings and the running records which show the history of robot's collision with wall, the detected information about the running circumstances can be integrated with the information indicated beforehand on the route map.(3) The results mentioned above make it possible to realize the mobile robot control by integration of the range sensor data and the touch sensor data.