Abstract
According to the classical walking robot inside the mine, a solution of the high-entropy controlling system is to be proposed. In the analog experimental environment joining the large data cloud computation, through the infrared environmental information detected by the multi-sensors to track the target object, the goal of no-track and crawler-type walking are achieved. Optimizing the walking of the robots has been verified through the Matalab emulation experiment and the real substantive testing, which establishes a solid foundation for the loading system inside the mine.