Developing A Route Navigation System Using Genetic Algorithm

Rout planning system is one of the interesting problems for a car navigation system. There could be many possible routes for a driver when a set of origin-destination pair is given. This paper addresses the problem of selecting route to a given start and destination on an actual map under user defined environment (static environment). The proposed solution using genetic algorithm (GA) with variable length chromosomes, i.e. chromosomes size could be dynamic due to path length (number of cities between source and destination). The population size of routes is generated with respect to the chromosome size. Two selection methods are used and compared: tournament selection method and a new suggested selection method based on selecting the best next neighbor. Number of crossover points (1-point and 2- point partitions), and different mutation points (1- point and 2-point partitions) used and tested to study and evaluate the effect and the behavior of these operators on the rout navigation system. Experiments showed that using the suggested route navigation system is quite efficient in calculating an optimal route incase the chromosome length equals the total number of cities in the map (i.e. the database), while when the chromosome size is less than the total number of cities, semi-optimal route is found.