Control system of field pipeline detection robot

The invention relates to the technical field of pipeline detection robots, and discloses a control system of a field pipeline detection robot, and the system comprises a fuzzy neural network obstacle avoidance algorithm based on pytorch deep learning, and the risk degree of collision can be calculated through the algorithm. In the obstacle avoidance process of the robot, only the relation between the distance between the robot and the obstacle and the direction of the relative speed and the collision risk degree, namely the space collision risk degree, is discussed. And the shortest distance and the relative speed orientation of the robot relative to the obstacle are used as parameters. The collision risk degree is calculated by adopting a fuzzy neural network obstacle avoidance algorithm based on pytorch deep learning, movement of a machine is decided through series operation and data transmission, working platform man-machine interaction is completed through a wireless WiFi module, the situation of obstacle