Synergistic approaches to mobile intelligent transportation systems considering low penetration rate

This paper investigates the effect of low penetration rate on mobile phone-based traffic state estimation (M-TES) models. Synergistic approaches, including an appropriate genetic algorithm (GA) based velocity-density estimation model and a notable artificial neural network (ANN) based prediction method for unacceptably low penetration rate, are proposed. The GA-based traffic state estimation model not only improves the effectiveness but also reduces the critical penetration rate required in the M-TES model. When the critical penetration rate is reduced the error-tolerance and the scalability of the estimation model can be significantly improved. The ANN-based prediction approach is introduced to overcome the weakness remaining in the GA-based traffic state estimation model when the penetration rate becomes unacceptably low or unknown. In addition, the effect of related road segments on the prediction effectiveness is thoroughly discussed. This work, therefore, provides practical instructions in narrowing the search space for finding prediction rules of the ANN model, thus improving the computational performance without compromising the prediction accuracy. The experimental evaluations confirm the effectiveness as well as the robustness of the proposed approaches. As a result, this research contributes to accelerating the realization of mobile phone-based intelligent transportation systems (M-ITS) or, of the M-TES systems in specific, since the essential issue of low penetration rate has been solved.