Research On Task Planning Of Simulation Service Robot

Since robot appeared in the1960s, robotics has experienced decades of rapid development, the use of robot begins to penetrate to the military, medical, services, etc., from the traditional industrial areas. As a young member of the family of robot, the service robot is a semi-autonomous or autonomous working machine, and it can perform various service works that benefit humanity, and improve the quality of production and life of mankind. With service robot entering into the family life in recent years, the market of robot is expanding. There are more countries have great attention to the research and development of robotics. The service robot has a bright future.During the development of service robots, robot task planning as a key technology is a hot and difficult research. In this paper, we use the simulation model given in the robot competition to describe the home service robot task planning model, and treat the simulation platform as a research platform to research the method of robot task planning. At the same time we also test the effect of each method of task planning in the simulation system. The main work of this paper is as follows:1. Transform the home service robot task planning problem into the representation of the answer set programming, and solves programming problems by iclingo planner. Answer set programming is a highly effective intelligent planning technique which has the ability of knowledge representation and non-monotonic reasoning. We can use planner to solve problem after we describe the planning problem by answer set programming’s specific language, without designing complex algorithm. The experimental analysis shows that the answer set programming can efficiently solve out the optimal solution of planning problem, but with expanding of the scale of problem, the time to get solution is also growing.2. Design a specific strategy for solving task planning problem in connection with home service robot task planning models. This strategy divides the state space of the robot planning after detailed analysis of the problem model, and describes how to select an action in each kind of state. The experimental analysis shows that the strategy can always give a solution for each task planning problem in a short time, but it can not guarantee that the solution solving out is the optimal solution. In addition, the scalability and ability of processing constraints are not as good as answer set programming, because the strategy is a hard programming method.3. In order to improve the efficiency and scalability of planning process, this paper applies the IDA*algorithm to task planning. IDA*algorithm is an iterative deepening A*algorithm, which solves the problem of high space complexity of A*algorithm, and is a highly effective heuristic search algorithm. The core of the heuristic search algorithm is valuation function, which are usually designed for a specific problem. In this paper, we design a valuation function for robot task planning problem. The experimental analysis shows that this valuation function has a good effect and short planning time.