Study On The Adaptive Route Planning System Of Intelligent In-Vehicle Information Equipment

Traffic problems such as traffic congestion have been increasingly prominent, which have seriously worse influence on the peoples' normal work &life and also have prevented the national economics from further development. Faced with this situation, Dynamic Vehicle Guidance System (DVGS) was proposed to be an effective approach to solve these problems. In-Vehicle Information Equipment (IVIE) is a key part of DVGS. It has experienced three eras of transformation and development. However, with the further penetration of IVIE, the existing IVIE has presented some shortcomings. Therefore, relying on the national science & technology tackling project "Development of In-Vehicle Information Equipment" and the 985 project "Development of Intelligent In-Vehicle Information Equipment", this paper carries on some research on the problems, so as to develop an Intelligent In-Vehicle Information Equipment (IIVIE) to satisfy all kinds of drivers' requirements. Thus, it makes the DVGS has effect on alleviating traffic congestion.The first chapter mainly introduces the importance and significance of the study. First of all, it introduces the definition, architecture,classification and function of DVGS.Then, it introduces the developing status of IVEE,which plays an important role in DVGS.It analyzes the market demand and the shortcomings of existing IVIE, (for example,the existing ones are based on static single criteria;the guidance route could not change according to the different situation;it could not supply the drivers with some personalized services such as recognizing the drivers' preferences;when the market penetration rate exceeds 30%~40%, distributed type IVIE may cause Braess Paradox).It also proposes that the IIVIE is the future developing direction and analyzes its significance and feasibility. Adaptive Route Planning System (ARPS), which is the key function of IIVIE is proposed. At last, it presents the main research content and arrangement of the whole paper.The second chapter begins with the analysis of the IIVIE's function. It proposes the scheme of IIVIE's software function and hardware integration. And also give a particular introduction of the ARPS' architecture design scheme, eleven modules (such as the user database management module, multi-route planning module, route parameter display module, route display module, route selection module, route modification module, preferences learning module, weight normalization module, multi-objectiveive resistance function computing module, adaptive route planning based on drivers' preferences module) and the developing flow charts.The third chapter summarizes the existing conclusion of the drivers' route selection behavior.lt analyzes the influence factors on drivers' behavior and also some problems caused by drivers' certain behaviors, such as information overmuch, ultra reaction and centralization reaction.Then, it analyzes the formation of Braess paradox and proposes six kinds of feasible solving strategies. At last, it enumerates some research method of the drivers' preference and designs a questionnaire survey of the drivers' route selection behavior in Changchun city. Based on the actual questionnaire data, it analyzes the different route selection preferences of different drivers.lt also comes a conclusion that the drivers' route selection preferences have spatio-temporal characteristics. That is to say, drivers will have different preferences according to different destinations and different journey time (weekday and holidays).The fourth chapter analyzes the disadvantages of single objective resistance function. And it designs a kind of multi-objective resistance function on the basis of summarizing the researches of multi-objective resistance function in and out of China. Then, according to the results of principal components analysis, hierarchical cluster analysis and weighted grading analysis, this paper selects the highest grade from the five criteria clusters. So the "Travel time","Travel cost",'Travel distance" and "Road classifications" are chosen as the four criteria for the multi-objective resistance function. After that, a driver's preference learning algorithm model based on differential perceptron is proposed to decide all the weight of the criteria and a weight normalization method is proposed to make the sum of all the weight equals 1. Finally, it proposes the realization method of all the models.The fifth chapter firstly introduces the advantages and disadvantages of several classical shortest route algorithm, including the Dijkstra algorithm, Bellman-Ford-Moore algorithm, Floyd algorithm, heuristic A* algorithm, heuristic B* algorithm, heuristic D* algorithm. Then, it summarizes the improved algorithm such as improved data structure, bidirectional search, layered search and some artificial intelligent methods. Then, on basis of these algorithms, it proposes an adaptive route planning algorithm based on drivers' preferences. This algorithm applies improved data structure, layered searching to enhance the speed of searching speed. The kernel searching algorithm adopts the heuristic B* algorithm. Meanwhile, it designs a heuristic function based on drivers' multi-objectiveive resistance function and brings forward its realization steps.The sixth chapter mainly introduces the adaptive route planning system. Firstly, it summarizes some human-machine interfaces design schemes and interaction schemes. Then, under the leadership of these rules, this paper has developed the adaptive route planning system VI .0 with the help of Visual C++6.0, MapX 4.0and Access database. This system could offer some functions, such as user recognition, user management, multi-route planning, map operation, drivers' preference learning, modifying routes, route planning based on drivers' preferences and help. Particularly, it also introduces the implement steps for the featured function of the system—drivers' preferences learning. At last, this paper selects 20 drivers as the experimentalists to validate the two algorithms and the feasibility of the whole system. It comes to a conclusion that the system this paper has developed is effective and different drivers will have different preferences. If we offer the optimum route based on their preferences, these routes will be different from each other. In this way, it will alleviate the Braess paradox to some extent.The seventh chapter summarizes the achievements and the shortcomings of this paper and also gives some prospective advises of the IVIE.The achievement includes:? Completing the software function design and the hardware design;? Completing the architecture design,modules design and developing flow design of the adaptive route planning system;? Completing the formation analysis of the Braess paradox and proposes six kinds of solving strategies;? Completing the questionnaires survey of drivers in Changchun city and analyzes the preferences of them;? Using principal component analysis and weighted grading analysis to decide the four criteria within the multi-objective resistance function model;? Proposing the drivers' preferences learning algorithm model and the realization method based on differential perceptron;? Proposing the adaptive route planning algorithm model and the realization method based on the drivers' preferences;? Developing the adaptive route planning system with the help of Visual C++, MapX 4.0 and Access database.