A Novel GA-BP Based Bidding Prediction Algorithm for Contract Logistics of Road Freight Transportation

Abstract

In this paper, we propose a novel bidding prediction algorithm based on the hybrid GA (Genetic Algorithm) and BP (Back Propagation) model for contract logistics of road freight transportation. Seven factors of GDP, fuel price, market requirement, cargo weight, transportation distance, carrier, and truck specification are concerned as the major factors. According to the nonlinear nature of transportation cost and carriers' psychological price, the BP neural network is applied to fit the nonlinear features. By training the dataset, GA is good at predicting the short-term bidding prices. In addition, genetic algorithm is integrated into the GA model to improve the long-term searching performance. The hybrid GA-BP algorithm is experimented to be of a good predictor. This study gives implication to 3PL providers that mining previous records with artificial intelligent algorithm is effective and feasible in road freight transportation bidding managements.