Road Risk Modeling and Cloud-Aided Safety-Based Route Planning

Abstract

This paper presents a safety-based route planner that exploits vehicle-to-cloud-to-vehicle (V2C2V) connectivity. Time and road risk index (RRI) are considered as metrics to be balanced based on user preference. To evaluate road segment risk, a road and accident database from the highway safety information system is mined with a hybrid neural network model to predict RRI. Real-time factors such as time of day, day of the week, and weather are included as correction factors to the static RRI prediction. With real-time RRI and expected travel time, route planning is formulated as a multiobjective network flow problem and further reduced to a mixed-integer programming problem. A V2C2V implementation of our safety-based route planning approach is proposed to facilitate access to real-time information and computing resources. A real-world case study, route planning through the city of Columbus, Ohio, is presented. Several scenarios illustrate how the "best" route can be adjusted to favor time versus safety metrics.

A Framework for Traffic Collision Prediction Using Historical Accident Information and Real-Time Sensor Data: A Case Study for the City of Ottawa

“…A periodical re-evaluation of these trends and their customization for specific regions can be included in an intelligent system to improve the predictions of collisions by either considering them as input variables or by weighing stronger prone days, times and months [25]. This is an aspect that will be studied and incorporated in this thesis work.…”

“…Hybrid neural network solutions, such as adaptive neuro-fuzzy techniques (e.g. ANFIS), are proposed by [35], [36], [25] and Grey-neural network hybrids by [37], [38].…”

“…Li et al [25] use a hybrid neural network model to predict the road risk index (RRI). The raw data is split into 3 clusters using fuzzy C-means clustering and, for each cluster, a separate neural network is developed to predict a different crash rating:…”

“…Another often used method is fuzzy techniques and their variants, including fuzzy logic [40], [41], fuzzy regression [42], fuzzy clustering [43], [25], evolutionary fuzzy rules [44], fuzzy rules and genetic programming [45], fuzzy granular decision trees [46] and hybrid neuro-fuzzy approaches (ANFIS) [36], [25], [35].…”

“…To show this, let's assume that the number of TN is almost equivalent to TP, and because both classes are similar real positives are also almost equivalent to real negatives, the number of FP and FN are also equivalent as shown in equations (25) and (26).…”