Laser slam-based autonomous navigation for fire patrol robots

Abstract

The ship structure is complex, the area is closed, and there are many flammable and explosive materials, so there is a great potential risk of fire. Once a fire occurs, the economic loss is serious and the risk factor of manual fire fighting is large. Therefore, it is significant to reduce the risk factor and grasp the fire information in time. This wen designs and develops a ship fire patrol robot. The article gives a detailed description of the robot design overview, system composition and function description and completion and performance parameters, and gives a brief overview of what can be expanded in the future. The overall architecture of the project is divided into a hardware part and an algorithm part. The hardware part uses the stm32F407 as the main control board, and its main control system contains the mechanical device, WIFI digital transmission module, etc. The auxiliary part uses the Raspberry Pi as the auxiliary control board. In the algorithm part, the project is divided into a planning and decision layer, a motion execution layer, a sensing layer and an application layer. The main technical features of the project include the use of McNamee wheel, the use of Robot-Link V4.0 WIFI digital transmission module, the use of deep learning based regression methods, etc., and has optimized innovation in remote monitoring interface, autonomous navigation autonomous patrol and regulation.