Derivation Models to Maximize Railway Track Operation in Egyptian National Railway (ENR)

Abstract

Railway capacity improvement is considered the base of railway network developments; so, it needs extensive planning to optimize the limited resources. Due to the continuous increase in the number of population in Egypt, it is expected that the railway capacity will be increased to accommodate the traffic demand. Decision-makers do not have a tool for determining how to increase the current capacity reaching the maximum possible capacity under the current operating conditions. The present thesis estimated the network performance by evaluating the Egyptian railway network based on the official timetables data then determining the practical railway capacity using artificial neural network techniques. Based on the neural network parametric model, the capacity of 56 out of 95 studied ENR network links can be increased under the current operating conditions. The practical capacity of the remaining 40 links is less than the used capacity. This means that these links are over-saturated and should be improved to safely accommodate the official number of trains. It was noticed that the largest zones with linkages that could be improved were in Middle and Southern zones, with an improvement ranging between (9 to 56%) of the used capacity. Also, the maximum percentage increase in used capacity was in the West Delta and reaches 83%. The proposed models investigated many factors regarding their impact and assigned them according to their importance as follows: the longest block section, type of signal system (mechanical or electrical), operating speed (for passenger and freight trains), track type (single or double), number of blocks, number of level crossings, and traffic composition ratio respectively. This arrangement is suitable for the operating conditions in Egypt. Analytical models have been derived to calculate the railway capacity in terms of the current operation conditions (track types and signaling and interlocking systems) based on the most significant factors obtained from the neural network model which are considered the longest block section, operating speed of passenger and freight trains. Then, iteration techniques were applied to obtain the optimum value of capacity effective factors reaching the maximum railway capacity. Thus, more than 65% of 56 links that can increase their capacity can reach maximum capacity under some changes through decreasing the block length and or increasing operating speeds of passenger and freight trains within the current operating conditions. Decreasing the block length must be studied separately in detail that is out of scope of this research. The Egyptian railway network has been studied as links; further juctions capacity study must be also studied in detail because it was also out of scope of this research. The links that can be maximized are concentrated in the West Delta zones and their capacities can be improved about 88% of the used capacity.