Hierachical model-algorithmic complex of combined planning and proactive control regarding the example of railway train group coordination

Abstract

Aim. At present there are quite a few systems that automate the control of a train or group of trains in order to reduce manual control and create an optimal schedule within the railway network. However, those systems are mostly based on rules of thumb and lack universality, so they have a fairly narrow scope of application. The original software and algorithmic complex considered in this article allows to solve this problem of multi-level planning, optimizes the capacity of the railway network, and minimizes violations of the due arrival times. Materials and methods. An original multi-model software and algorithmic complex, divided into several layers is used to solve the described problem of multi-level planning. At the lower level of the hierarchy, the movement of an individual train is planned, the next level is devoted to regarding the processes of passing a particular section of the network to technological works in terms of network planning, and at the upper level evaluation and optimization of integrated quality indicators for a whole group of trains, as well as evaluating and adding stability reserves to the built plan take place. Results. An original method and a software-algorithmic complex based on it were developed. This system solves the problem of planning and coordinating a group of trains within a single railway network in order to optimize its capacity and minimize violations of arrival deadlines. Along with this, the developed complex solves the problem of proactive real-time train control, and can be used as a decision support system. Conslusions. The proposed methodic, using fundamental approaches, can signifi cantly reduce the complexity of the original multidimensional computational problem by solving it by a hierarchy of models and reducing the original task to a network planning problem. The developed softwarealgorithmic complex solves a wide range of tasks in terms of train group movement planning and coordination, and has a high versatility. Therefore, the chosen approach has good prospects, but it requires further testing and application for real transport systems to prove its effi ciency.