Abstract
One of the essential attributes of a cleaning robot is to achieve complete area coverage. Current commercial indoor cleaning robots have fixed morphology and are restricted to clean only specific areas in a house. The results of maximum area coverage are sub-optimal in this case. Self-reconfigurable tiling robots are innovative solutions for such a coverage problem. These new kinds of robots can be deployed in the cases of cleaning, painting, maintenance, and inspection, which require complete area coverage. Tiling robots’ objective is to cover the entire area by reconfiguring to different shapes as per the area requirement. In this context, it is vital to have a framework that enables the robot to maximize the area coverage while minimizing energy consumption. That means it is necessary for the robot to cover the maximum area with the least number of shape reconfigurations possible. Conventional tiling theory based area coverage methods are usually tedious. A layout is initially decomposed into tilesets based on the tiling theory. Once the tilesets are generated, then various algorithms like zigzag planner, spiral planner among others are used to generate paths by connecting the tilesets ensuring complete area coverage. There is another method where by considering this problem as a travelling salesman problem(TSP) which is solved by traditional Genetic Algorithm(GA) and Ant-Colony Optimization(ACO) algorithm. Even with these methods, as the layout complexity increases the computational requirements increases and the tilesets that are generated are also not often optimal. The current thesis attempts to address this problem by proposing a complete area coverage planning module for the modified hTrihex, a honeycomb-shaped tiling robot, based on the deep reinforcement learning technique. This framework simultaneously generates the tiling shapes and the trajectory with minimum overall cost. In this regard, a Convolutional Neural Network(CNN) with Long Short Term Memory (LSTM) layers are trained using the Actor-Critic Experience Replay (ACER) reinforcement learning algorithm. The results obtained shows that the proposed method gives a better cost-weight path in a quicker time.