The Study Of Search Path Planning For Intelligent Vehicle

After many discussions with the supervisor and the same group of members,thesmart car tracing studies planning as a subject of study in my. Although the researchon Intelligent Vehicle planning, diagnosis, but I to path planning as the researchobject,not in the study of diagnostic study this subject category.Tracing planning membership intelligent planning research category, the researchof intelligent planning in other words the description logic (DL),planning andprogramming language description of related algorithms. Planning representationlanguage mainly centered in the state space description language STRIPS (STRIPSmethod is based on “Stanford Research Institute Planning System” this academicinstitution named after.),Action-centered description language ADL (ActionDescription Language) and planning domain definition language PDDL (PlanningDomain Definition Language).Research on how to solve the planning problem is nothing more than the variousplanning algorithms are studied, the objective is to achieve: When the state spaceinformation has the problem can be described, and whether it is original or collectionof information for the target state and the corresponding actions are, we can useefficient inference, finding a path planning scheme between the original state and thetarget state, in other words, is to give an effective action sequences, and from theoriginal state through the target state after the execution of the action sequence.Through the comparative study of state space search method, partial orderplanning algorithm, hierarchical task network planning (HTN), planning (Graphplan),planning method of heuristic programming method, timing and SAT (Satisfiability) algorithm, SAT algorithm as the main algorithm was chosen to experiment theresearch. Based on the comparative study of various planning method can meet thedegree of attention conversion by the planning algorithm based on the maximum, thealgorithm in use is to put the planning problem is transformed into an SAT relatedproblems，and then solve. In this stage, appearing in the SAT solver to solve is themost effective but also one of the most advanced tools considered, which is basedCNF (Conjunctive Normal Form) to be expressed in the form of questions.Application of SAT planning method can be summarized as follows:1, the known planning issues are converted into coded propositional formula;2, to the propositional variables in the formula assigned different values, andthen apply the formula to meet the judging process approach to determine, to observewhether these assignments so that the formula has to meet resistance;3, the assignment results from the decision process to meet identified in the planto extract the ideal solution.Path planning has developed into a comprehensive application of a variety ofmethods of planning, the paper in this regard only made a preliminary attempt, inexperimental studies testing process, the experimental effect is not very satisfactory.There is some intelligent vehicle path planning problems that require further study,such as the number of smart cars and coordinated planning, etc. These are moreadvanced planning system to be solved.