Research And Implementation Of Multi - Robot Cooperative Research Based On Behavior

This subject analyzes the "perception shortage" problem in the "smart city" and solves the problem with the ideas of using collaborative multi-mobile robots. This paper also designs a model and a fast method for processing image information, a route map model for robots, and a collaborative mechanism based on rank and authority. All these above make robots move to their destination conveniently, be any combination of different ones and complete the work collaboratively. In this way, people’s quality of life will be improved, and our city will be more modern and "wisdom".In analogy to the human brain, at the same time reference stage treatment, this paper uses a two level control structure which the motion controlling and the information processing are divided into two parts. No matter what kind of executing agencies the lower controller use, it can only do the behavior conversion and can effectively perform a behavior ordered by upper controller. According to the need of current state, the upper controller should use the correct sensing devices to acquire environment information, then decide behavior which will be executed by the lower controller. In this way, behavior decision can be separated from behavior decision execution, and be more versatile. This subject uses robots of Bioloid series and some other modules are installed on the robot to achieve the whole system.In an unknown environment, the robot needs to use a variety of effective means to explore, to safely and quickly reach the destination, and the requirements to robot are higher on various aspects. In the human living environment, there are some small independent areas which are relatively closed, such as residential areas, campus and so on. These regional environmental conditions are usually good. Using robots to patrol in it, will both save the labor cost, and achieves better patrol effect. Because the robot operates in the known environment, using a map to guide the robot, can greatly enhance the operating efficiency of the robot. Here designs and uses a route map model based on key point. It uses some real objects which have significant features as markers, and is convenient for robot path planning. And its complexity and data volume are below the traditional map model.Vision is an important way to perceiving the world and a huge amount of information is its remarkable characteristics. It is very inconvenient to process fast, and complete access to all information is almost impossible. Conventional image information processing stresses the point. It makes the treatment too detailed and complicated. This subject designs a method which stresses the object, and it helps robot tell markers apart from complex scenes, then the robot can know its location. This method uses CMPW model for color images to process RGB images and also binocular vision so the robot can find markers than move freely.This subject compares the pros and cons of the single robot operation and multi-robot cooperative operation and sets forth the necessity and feasibility of the multi-robot cooperative operation. Here puts forward a multi-robot cooperation mechanism based on rank and authority. In general situations, different robot runs on the independent line, and keeps observing. Once an accident is found by a robot, the robot will call up a temporary team to handle the accident collaboratively. After solving the accident, all the members of the temporary team will returned to their original routes.