Abstract
The genetic algorithm (GA) is a stochastic searching, learning and optimizing method for computer simulation. Utilizing GA, we have developed a basic simulator to find an optimum transportation route which can reduce transport cost of agricultural products. The simulator is coded by C language based on ANSI. It can find an optimum transportation route, given several parameters which are traffic volume, geographical information of passing through points, performance of delivery vehicle, etc. Two pairs of simulation experiments were designed and carried out : 1) traveling with empty load and delivery, 2) picking up cargo and delivering. As a result, the average fuel cost of the transportation route found by GA is considerably lower than the one obtained by random searching (RS) method in each simulation. On the other hand, the searched optimum transportation routes did not correspond between traveling picking up agricultural products and delivering, under the condition that each cargo quantity was same. These results lead us to the conclusion that our simulator is useful for searching an optimum transportation route and that the big cost cut in transportation of agricultural products is realized by combining GA and concrete transportation data about target agricultural products.
