Development of optimal and cost effective bus scheduling using genetics algorithm

Abstract

Most higher institutions in Nigeria operates intra and inter campus transportation, but lack proper planning on movement schedule which contributes significantly to poor academic performance of students as it causes great fatigue due to long queue at the parks and consequently, resulting to losses of revenue to bus management. A met-heuristic algorithm, G.A and L.P model were used to optimize bus scheduling for efficient transportation. To achieve this, travel demands for peak and off-peak seasons at the two campuses of FUTMinna, Nigeria were obtained using 4 numbers of CCTV cameras located at strategic positions. Data were analyzed using the design travel times of 40, 50 and 60 minutes considering traffic and road conditions using 19 numbers of 18 seater bus, 11 numbers of 35 seater bus and 15 numbers of 60seater bus capacities buses with15 minutes departure time. It was observed that with N100/head of revenue charges, a total of N185,000/day corresponding to 231 trips would be achieved during off-peak season to convey 2,562 students, and this amount would increase by 289% at peak season with very little or no delay to move 4,411 students travel demand.