Search

Over the past few years, several studies have considered the problem of Electric Vehicle Path Planning with intermediate recharge (EVPP-R) that consists of finding the shortest path between two given points by traveling through one or many charging stations, without exceeding the vehicle’s range. Unfortunately, the exact solution to this problem has a high computational cost. Therefore, speedup techniques are generally necessary (e.g., contraction hierarchies). In this paper, we propose and evaluate a new fast and intuitive graph clustering technique, which is applied on a real map with charging station data. We show that by grouping nearby stations, we can reduce the number of stations considered by a factor of 13 and increase the speed of computation by a factor of 35, while having a very limited trade-off increase, of less than \(1\%\), on the average journey duration time.

This paper presents a novel approach for trip reconstruction and transport mode detection. While traditional methods use a fixed GPS sampling rate, our proposed method uses a dynamic rate to avoid unnecessary sensing and waste of energy. We determine a time for each sampling that gives an interesting trade-off using a particle filter. Our approach uses as input a map, including transit network circuits and schedules, and produces as output the estimated road segments and transport modes used. The effectiveness of our approach is shown empirically using real map and transit network data. Our technique achieves an accuracy of 96.3% for a 15.0% energy consumption reduction (compared to the existing technique that has the closest accuracy) and an accuracy of 85.6% for a 56.0% energy consumption reduction.

One of the major types of cardiovascular diseases is Coronary Artery Disease (CAD). This study tackles the problem of CAD detection using a new accurate hybrid machine learning model. The proposed ensemble model combines several classical machine learning techniques. Our base algorithm is used with four different kernel functions (linear, polynomial, radial basis and sigmoid). The new model was applied to analyze the well-known Cleveland CAD dataset from the UCI repository. To improve the performance of the model, we first selected the most important features of this dataset using a genetic search algorithm. Second, we applied a multi-level filtering technique to balance the data using the ClassBalancer and Resample methods. Our model provided the average CAD prediction accuracy of 98.34% for the Cleveland data (the average was taken over the four kernel functions).