Intelligent prediction of traffic volume distribution

Traffic issues become one of the most important problems these days because of the life style of human. Therefore, paying more attention to this field seems to be essential. Distributions of traffic volume on urban roads is often described by some statistical models, which are based on probability and are suitable for ideal physical and environmental conditions. However a noticeable point is that they don't have the capability of working under complicated situations, all due to their mathematical constraints. In this paper the utilization of Neural Networks in the field of predicting traffic distribution for Petaling Jaya-an area in south west of Kuala Lumpur- is being discussed, which is applicable to a wide range of traffic situations and can play an important role in responsive urban traffic control system. A neural network-based system approach is implemented to establish an adaptive model for simulating traffic volume distribution and consequently its prediction. It has been found that Neural Networks can act strongly in the case of traffic prediction