Electric Control System of Smart City Marine Garbage Cleaning Robot Based on Deep Learning

Abstract The problem of water resource security is closely related to human life. How to maintain the safety of water resources and timely clean up water pollutants has become the focus of human attention at this stage. The ocean is an important source of water resources, and marine garbage cleaning is very important. This study will analyze and design an electronic control system of a garbage cleaning robot by a deep learning algorithm. The experimental results show that the access of maximum power point tracking equipment is conducive to maintaining the functional stability of the robot power supply system in different environments; the solar cell maintains the maximum output power and prolongs the robot’s endurance time; and the target recognition algorithm based on deep learning can achieve 95 % accuracy rate, 1.25 % false alarm rate, and 5 % false alarm rate respectively, which has high reliability. When people identify and locate the three places and collect garbage, the azimuth error is 8.4 %, 4.6 %, and 3.0 %, and the distance error is 5.6 %, 4.9 %, and 11.1 %, respectively. In conclusion, the electric control system can guide the robot to complete the task of marine garbage cleaning. It is hoped that the research results can provide help for the development of a marine garbage cleaning robot.