The Research Of Heuristic Search Algorithm Based On Quotient Space Of Granular Computing

Searching technology is a basic and important research field of artificial intelligence(AI).AI deals with the information which is always non-confirmed,fuzzy,non-completed and massive. So generally it is not sure for us to know the way for problem solving. Instead, we should search the problem space to find the right answers.When it is short of neighborhood knowledge and the scale of problem is not so great,we can make use of blind search strategies such as breadth-first search,depth-first search etc.In these cases, the cost of time is always acceptable, so blind search technology is widely adopted in small-scale expert systems.As a real artificial intelligent system, its problem-solving space is very large, and every node in the space denotes a subgoal of solutions.For every subgoal, there are some solving methods to select. So we can't practically solve those problems if we use the blind search strategy in many cases. Heuristic search use the experienced and intuitive knowledge to control the direction and path of search, which overcomes the blindness of searching to a certain extent and enhances the solving efficiency a lot. But the present various heuristic-based search algorithm has basically not overcome the calculation of the index explosive issue.Granular computing is a branch of soft computing and it is also a emerging conceptual and computing normal form of information processing. Just as a great umbrella, it may be regarded as a label of theories,methodologies,techniques and tools that make use of granules. Granular computing is a super set of the theory of fuzzy information granulation, rough set theory, the theory of the quotient space and interval computing etc. It plays an important role in information processing for fuzzy,uncertainty,partial truth,mass and granular computing is becoming one of the main study stream in A I.This paper proposed a heuristic search algorithm based on quotient space .In additional, it discussed and researched the granular principles of the process of heuristic search in depth. This algorithm makes use of hierarchical method to divide the solving space in a coarse granule firstly, and then judge which granule contains the solution by heuristic information. If so, divide the granule in a smaller granule, otherwise give up to deal with the granule. Repeat this process untilfinding the right solution. Based on this hierarchical thinking, the algorithm reduced a lot of unnecessary search and improved the efficiency of problem solving which well simulated the human analysis and problem-solving process. In other words, it is able to observe and analyze the same issue in different granules and synthesis various information of each granules of different size to get the solution. In the end of this paper, it verified the algorithm based on granular computing, the experiment proves its validity in edge information extraction of the pictures. It obtained better extraction effect than the traditional method.