Portal de Periódicos da CAPES

Invited PaperHigh-speed, High-density Train Allocation

1970; WIT Press; Volume: 7; Linguagem: Inglês

10.2495/cr940022

1746-4498

Satoru Sone,

Railway Systems and Energy Efficiency

In order to realise higher capacity and shorter passengers' time with restricted track capacity, new systematic methods of making and evaluating many good new train stopping patterns have been discussed. This paper includes three major proposals; (1) a new trains' loci chart has been established, whose time axis is taken as normalised time difference between stopping and passing trains, the expression of which is more easily understandable for us and more easily treatable for a computer than ordinary expression, (2) performance of zonal separation train scheduling, which is known most suitable for typical commuter lines connecting city centre and suburban area, has been improved by selective additional stops, and (3) a method of finding the best stopping pattern out of all possible patterns which are made systematically by a computer with a section of a line, number of trains, cycle time and number of trains serving each station given. Combination of the above proposals can realise theoretically best train scheduling as far as the data given and assumptions taken are reasonable. INTRODUCTION Urban and intercity rail services are very important from social viewpoints such as traffic safety and environmental friendliness, etc. They are, however, not so attractive from users' viewpoints because of poor train scheduling. Especially, in densely populated area with insufficient social investment, trains are congested and passengers' time is too long. If number of tracks could be increased and each kind of trains such as fast, semi-fast and local, could run with its own track, the problems would be solved, but it is very difficult to construct additional tracks in densely populated area. Under these circumstances proposals given in this paper would help railways receive much better social acceptance through improved passengers' services. Transactions on the Built Environment vol 7, © 1994 WIT Press, www.witpress.com, ISSN 1743-3509 12 Railway Operations The basic adea of the improvement is reduction of number of stops per each train; typically, 1 minute time is consumed at each intermediate stop with time loss of 15 seconds for deceleration, 25 seconds for station stop and 20 seconds for acceleration, and minimum practical train headway is 2 minutes for two consecutive stopping trains but less than one minute for two consecutive passing trains. The problem is the fact that reduction of number of stops does not automatically improve the situation; it may be easily understandable that if number of stops is unreasonably adopted, some passengers have to wait long until a later train serving their destination, or have to change trains which had not be necessary, or even they can never be reached at their destination, and passing trains can not always run faster because of the movement of preceding trains. This paper deals with how to determine intermediate stops of a group of trains in order to minimise passengers' time and maximise train capacity, and proposes the methods of finding the best solution out of all possible solutions both made by a computer. 1. USEFUL CHART OF EVALUATING REASONABLE TRAIN STOPPING PATTERNS AND ITS EXPRESSION FOR A COMPUTER USING BINARY NUMBER In order to realise fast and frequent train scheduling, it is very important to express time difference required for relevant trains as purely as possible. From this respect, train loci in distance time surface shown in such a way that time of each train is shown by actual time minus standard time of the fastest train was proposed by the author in theprevious paper[l]. In this paper, time difference is made at station only by whether the train stops or passes, and because actual distance or time required between adjacent stations has little meaning, distance axis is taken as order of the stations only. Practical train headways between two consecutive trains are fortunately close to the time difference above for passing trains and two times that for stopping trains. So, if time is expressed as normalised by the difference between stopping and passing trains, which is typically 1 minute, we can draw the train loci on the general section paper rather than specially prepared time distance chart. The fact that train loci exist only on the line of the general section paper or a shortest 45 degree line connecting diagonal points means that the loci are easily expreesed by 1's and O's on the computer as shown in Table 1. Table 1 Expression of Figure 2a for computer 100000000001 1 1 1 1 stopping 1 0 0 0 0 0 0 1 1 1 1 1 0 0 0 0 pattern 10001 1 1 100000000 1 1 1 1 100000000000 010000000000000 operating 010000000000000 pattern 010000000000000 000000000000000 Figure 1 shows typical train loci without additional loss times. If a passing train follows a stopping train with a relatively short headway, it can not run at normal Transactions on the Built Environment vol 7, © 1994 WIT Press, www.witpress.com, ISSN 1743-3509 Railway Operations 13 speed, i.e. an additional time loss takes place. In this case the train loci can be expressed as in Figure 2a as far as possibility of dense train allocation is concerned. In practical application, reduction of running speed may better be put at other places such as in Figure 2b or 2c, by some reason such as energy saving, reducing running noise problem, increasing chance of opening level crossing barrier, etc. But this modification can be made at a later step of making train scheduling than discussion in this paper.