Uncertain Penetration Rate Issues in Mobile Intelligent Transportation Systems

This paper thoroughly discusses the essential issues remaining in mobile phone technologies which impede the realization of mobile phone based traffic state estimation systems (M-TESs). Concretely, the inherent issues in mobile phone based applications, namely low and uncertain penetration rate issues, which affect the M-TES’s effectiveness, are resolved. A unique GA-based velocity-density estimation mechanism is proposed to improve the traffic state estimation accuracy when the penetration rate is low but still be relevant. A novel ANN-based prediction model is proposed to cope with unacceptably low and uncertain penetration rate issues. Moreover, a reasonable selection method is proposed aiming at selecting an appropriate traffic state estimation model without the actual penetration rate information. Experimental evaluations reveal the effectiveness as well as the robustness of the proposed solutions.