To plan and control train traffic constitute very complex tasks. The human being, in the role of train traffic dispatcher, must be able to take fast and correct decisions and effectuate them efficiently. These operators must then also be supported by efficient decision support and control systems. In order to study the involved questions a train traffic simulator (TOPSim) has been developed at Uppsala University in Sweden. With this TOPSim simulator it is possible for a human operator to in real-time control the movements of the simulated trains running on a model of an arbitrary railway line. This paper describes the results of our experiments using the first version of the TOPSim simulator, with the focus on questions concerning the computerisation of the re-planning process. Professional train dispatchers have been participating in the experiments presented here. Also the questions of using a decision support tool based on a traditional train traffic simulation program, in the train dispatching process are discussed. These questions have some interesting implications on the traditional use of simulation in the long term planning process.

Manniska-maskininteraktion och granssnittsutformning i samband med tagtrafikstyrning : Rapport fran ett forskningsprojekt

All railway companies try to attain more regular and reliable train services, in order to be able to offer their customers a higher reliability and flexibility. The present, expensive, strategy is increasing ”physical” capacity. A more cost effective alternative is to first, as far as possible, improve quality in the train control process itself. The most important factor in this process is the train dispatcher, who from a control centre supervise (monitor and control) the movements of trains. The dispatchers chances to, in an active way, plan and control the train traffic is often crucial for how different disturbances effect train delays. Within this research area, one focus has been on development of algorithms and systems, which support optimal computer-aided train dispatching in real-time (CATD-RT). We present a brief overview of existing knowledge and research and some important findings from a survey based on a set of interviews with dispatchers and train traffic experts. Also indicated are further studies necessary in order to, at the end, be able to successfully implement a system of this kind.

Summary The rapid technical development of train traffic systems, together with increasing control demands, makes it necessary to develop new support systems for the train dispatchers. The dispatchers' support systems must have an appropriate functionality and a high usability. In order to design such control systems, we need a model of the complete train traffic system. This model must fulfil a set of different requirements. Main requirements concern descriptions of the relations between the goals for the planning and control of the train traffic system, i.e. the control objectives, and the strategies and performance of the operators. The model must also describe the technical support systems, the information systems and the dispatchers' use of these. A set of observation interviews have been performed, documented and analysed according to the described model structure. The analysis has resulted in a number of important conclusions for the design of a new generation of control support systems and of the operators' interfaces to these systems.