Multi-fire inspection collaborative robot system based on integral reinforcement learning

Abstract

The invention relates to a multi-fire inspection collaborative robot system based on integral reinforcement learning, which belongs to the field of robots. The system includes a hardware layer, an interaction layer, a perception layer, and a control layer; the hardware layer uses DSP as a controller, and sends the data collected by the odometer and gyroscope into the DSP for processing, and calculates the robot in real time. position in . The speed command is sent to the DSP through the upper computer, and the DSP will obtain the speed information code to control the operation of the servo motor; the fire inspection robot uses a crawler drive; when the mechanical arm needs to move, the ros system in the upper computer passes through In moveit! The platform plans the motion trajectory of the target point where the robot arm will move, discretizes the planned motion trajectory and sends it to the DSP. The angular velocity and acceleration of each axis obtained by the DSP controls the movement of the servo motor of the robot arm to reach the target point. .

Description

technical field [0001] The invention belongs to the field of robots, and relates to a multi-fire inspection collaborative robot system based on integral reinforcement learning. Background technique [0002] At present, the main structure of the common fire inspection robot is: wheel drive is used in the drive; flame detectors and temperature sensors are installed around the robot to facilitate fire detection; the robot is equipped with a camera in front of the robot to transmit the inspection screen through the wireless module to the control room; a fixed but rotatable fire sprinkler is installed above the robot for external water pipes or small water pumps to extinguish the fire; in terms of robot control, with the development of multi-machine collaboration ideas and theories, in order to complete For the inspection of large areas, in order to improve the inspection efficiency and reduce the difficulty of inspection, multiple fire inspection robots are usually used to coope...

AI Technical Summary

Problems solved by technology

[0003] However, there are also many defects in the above-mentioned system. Firstly, the wheel drive makes the robot have poor passing performance on stairs and rough roads, and the flexibility of steering and rotation is not high enough; and the accuracy of flame detection by using flame detectors and temperature sensors and timeliness cannot be well guaranteed, and the range of flame detection is also small; secondly, after the flame is detected, it can only realize the alarm function and transmit the position of the ignition point and the image of the fire situation obtained through the camera to the fire control room , a small number of fire inspection robots can also cooperate with their own fire sprinklers to extinguish the fire point under the remote control of firefighters, but overall they lack flexibility and initiative in fire response; finally, multi-robot In terms of collaborative control, the centralized control method makes each individual robot have no ability to choose actions and coordinate with each other, which makes the inspection efficiency, robustness and scalability of the entire system poor, and each robot is patrolling. The optimal time and energy in the inspection process cannot be guaranteed, which will reduce the overall battery life and anti-disturbance ability to the outside world, and the autonomous control and intelligent level still need to be improved.

Images