An efficient algorithm for vehicle guidance combining Dijkstra and A∗ algorithm with fuzzy inference theory

Most of the studies on vehicle navigation are based on the use of an existing path-searching algorithm. However, finding the optimum path requires a lot of computation, which is quite limited on mobile devices. To improve the efficiency of vehicle navigation systems, we need to develop a suboptimum path-searching technique to save on computation for car navigation systems. In this study, fuzzy theory is applied to the path-searching algorithm. The drivers' collected expertise is collected into the fuzzy rule base system. Combining fuzzy inference theory with Dijkstra's algorithm and A∗ algorithm significantly reduces path computation time, and the result is more similar to drivers' actual driving habits. In the simulation in the northern area of Taiwan, the results of our experiments show our proposed method is either more efficient or has a shorter path than Dijkstra algorithm, A∗ algorithm and other algorithms.