论文部分内容阅读
对于使用移动机器人在风速/风向变化较大的气流环境中定位气体泄漏源的问题,我们建立了一个定位模型.模型的输入为机器人在定位过程中实时获取的多传感器信息(激光信息、视觉信息、气体浓度信息、风信息等),输出为相应的搜寻行为或策略,主要包括避障行为、随机搜寻、视觉搜寻、化学趋向性搜寻、风趋向性搜寻、路径规划和气体泄漏源定位等.利用矩阵的半张量积理论,我们确定了这个模型输入和输出之间的结构矩阵.根据多传感器的测量信息,结构矩阵产生相应的搜寻行为或策略,由动态机器人有效地完成,以确定气体源的位置.本方法的可靠性经过机器人实地实验得到验证.
For locating gas leakage sources using moving robots in air flow environments with large variations in wind speed / wind direction, we set up a localization model. The inputs to the model are multisensor information acquired in real time by the robot during positioning (laser information, visual information , Gas concentration information, wind information, etc.). The output is the corresponding search behavior or strategy, including obstacle avoidance behavior, random search, visual search, chemical trend search, wind trend search, path planning and gas leakage source positioning. Using the semi-tensor product of matrix theory, we determine the structural matrix between the input and the output of the model.According to the multi-sensor measurement information, the structural matrix generates the corresponding search behavior or strategy and is effectively completed by the dynamic robot to determine the gas The location of the source.The reliability of this method has been verified by robot experiments in the field.