یک سیستم پشتیبانی تصمیم گیری هستی شناسی محور برای طراحی با کارایی بالا و هزینه بهینه سازی فریم های پورتال پیچیده راه آهن

Electrification structures design for railway systems is a crucial and complex process, since it compounds plenty of infrastructure elements, design decisions, and calculation conditions. In this paper, an ontology-driven decision support system for designing complex railway portal frames is presented and developed. A knowledge-rules database has been also developed relying on experts knowledge and complying with railway standards. Our system outperforms the current portal frames design methods by decreasing construction time and costs. As a result, an intelligent computer-aided design tool is provided, thus facilitating the task of seeking for the optimal portal frame, which is geometrically and structurally feasible, and cost-effective.

Railway companies and engineers have been trying to enhance and to regulate their design methods for electrification infrastructure for many years now (AENOR, 2009, AREMA, 1998, AREMA, 2003, Brown, 1927, Bond and Harris, 2008 and UIC, 1981). Computers have played an important role on this effort, as railway and engineering companies have developed several tools along the time to automate the design process. As some examples, in 1964, Andrews (1964) presented a tool to calculate the behavior of an overhead catenary system for railway electrification, and in 1969, West (1969) presented a program to make optimized structural calculus of railway. However, the traditional design systems are not enough to achieve the best-suited designs, as it depends on the knowledge of several expert engineers that must work closely together with the railway company experts to design good structural components that are at the same time compliant with the railway company regulation and the body of national and international standards. Trying to cope with the problems, expert systems (Bonissone, 1983, Chang, 1988, Domeshek and Kolodner, 1992 and Fenves et al., 1989) were developed and commercialized in 1980s as a research tool designed to diagnose and to solve technical problems in the railway area. Expert systems rely on experts knowledge. Thus it is very important to organize and systematize the information to be able to include experts knowledge into the system. To achieve this goal, ontologies have been proposed as a way to systematize information in very different fields. An ontology seeks to provide a definitive and exhaustive classification of entities, including the types of relations by which entities are tied together. Gruber (1993) described the principles to design ontologies for knowledge sharing. Davies, Fensel, and van Harmelen (2001) shows several examples to manage knowledge based on ontologies. Abanda, Ng’ombe, Tah, and Keivani (2011) describes and ontology to solve the problem of land delivery in Zambia. Eden and Turner (2007) shows a possible ontology for computer programs and the problems for building such ontology. Ontologies of railways are not easy to find. A first attempt was presented in Bjorner (2004), where a proposal was made to create a Railway Domain to join experts and to create ontologies to classify railway objects, which was addressed as a grand challenge. Recently, Mohan and Arumugan (2011) has presented a railway ontology using Web ontology languages and Semantic Web Rule Language (SWRL) as a way to describe and share railway infrastructure information. However, despite the efforts (Baden, 2000 and Bailey and Smith, 1994), CAD tools used to design and calculate the railway overhead wire support structures still lack the vision of an integrated approach to the design problem. Thus, this process is usually divided in several steps (requirements, design, structural calculus, etc.), that may also involve several organizations, sometimes not well integrated. For example, designing with AUTOCAD or 3D CAD (Veerhoek, 2006) in one organization, and calculating with CYPE (CYPE, 2010) or CALPE (Benet, Cuartero, & Rojo, 2000) in another. This way of coping with the problem generates a catalogue of hurdles, including sometimes the exclusion of railway company experts from the process, which may lead to inadequate designs that are failure prone or difficult to maintain. Furthermore, it takes not less than 3 weeks of several engineers to have a design of a complex railway framework portal. Thus, it is important to have integrated CAD tools for railways, as proposed recently in other areas (Yang, Chang, Hu, & Zhang, 2010a), and one limited example for railway yard may be seen in Yang, Wangand, and Hehua (2010b). Applying artificial intelligence to railway tools is a major trend in the XXI century (Meng, 2010 and Sadeghi and Barati, 2010). There are examples for railway capacity planning (Yung-Cheng, Mei-Cheng, & Jyh-Cherng, 2010) and safety control systems (Yaroslavtsev & Levchenkov, 2011). However, AI has not been applied to the integral process of railway infrastructure design, and this is the major goal of our work. The global aim of the research shown in this paper is to provide an intelligent computer-aided design tool, named SIA, that helps railway infrastructure designers to design and calculate safer and more efficient complex structures for railway electrification systems, especially overhead wire support structures such as cantilevers and frame portals. The final goal of that tool is to help designers to build infrastructure for the electric railway transport that are compliant to the existing normative and safe from the circulation and structural point of view to avoid dangerous situations that arise currently due to mistakes in design. SIA follows a holistic approach to design, including facilities to automatically build from scratch detailed structures, including platforms, tracks, complex railway portals, catenaries, and electrification systems. Our tools provides an analysis of three-dimensional structures, detailed component calculations, election of minimum cost and minimum weight designs, and constructive plans adjusted to the millimeter of the optimal solution. The final result is the project documentation that can be provided to the building company. The vital factors to achieve such a tool and to ensure its future are its acceptability to designers, its usability, and, of course, its correctness and flexibility to include expert knowledge in the field. As argued in West (1969), those factors depend upon the relationship fostered among the promoter and the designers, so that the railway engineering designers can trust the results of the system. Thus, it is very important to present the tools to the users and to have significant examples to show them the feasibility and utility of the tools. It is also very important to be able to demonstrate to the railway authorities that the final designs respect the regulation and the national and/or international standards, as they will be accepted only if they are compliant with the normative. To demonstrate SIA versatility and usability, we show in this paper the process of designing and calculating a railway portal from scratch. The paper is organized as follows. Section 2 presents motivation and description of the railway portal design problem. In Section 3, the ontology used to represent knowledge in SIA is described. Section 4 shows the rule-based system that guides the design process itself depending on the ontology objects and their properties. Section 5 presents some results. And, finally, Section 6 shows the main conclusions of the paper.

In this paper, we have presented an intelligent computer-aided design tool, named SIA, that helps railway infrastructure designers to design and calculate safer and more efficient complex structures for railway electrification systems, especially overhead wire support structures such as cantilevers and frame portals. The final goal of that tool is to help designers to build infrastructure for the electric railway transport that are compliant with the existing normative and safe from the circulation and structural point of view to avoid dangerous situations that arise currently due to mistakes in design. SIA follows a holistic approach to design, including facilities to automatically build from scratch detailed structures, including platforms, tracks, complex railway portals, catenaries and electrification systems. Our tool delivers an analysis of three-dimensional structures, detailed component calculations, election of minimum cost and minimum weight designs, and constructive plans adjusted to the millimeter of the optimal solution. The final result is the project documentation that can be provided to the building company. The novelty of this research is the holistic approach of including all railway systems, with the intention of creating a knowledge system to incorporate automatically experts knowledge by using a rules engine, and the possibility of making automatically infrastructure design choosing optimal solutions, which will enhance the system efficiency and safety. SIA is a very complex system and it can be expanded with many functionalities. However, we have already implemented a prototype with includes a powerful functionality for tracks and electrification. ADIF experts are currently testing it, which is resulting in new enhancements and functionality. Its impact in the railway business may be important, as we do not know any tool similar to SIA, and it may allow to railway and engineering companies to make faster and safer designs, avoiding also expenses due to oversized systems or to failures due to erroneous designs. The link with the proposed challenges is clear. First, better design and calculations enhance safety because it avoids failures. Also the simulation of the interaction between pantograph and contact wire allows avoiding design failures in switches, where two contact wires are present and the pantograph needs to be always in contact with at least one of them and never go on any wire to avoid breaking it and the catenary. Those failures are serious because they are very expensive to be repaired and causes long traffic interruptions, thus SIA may help to make railway safer and easier to maintain. SIA can help to reduce design costs, by decreasing the design time and getting better solutions, by shortening the time for the infrastructure project to be available, and by reducing traffic interruptions due to failures. We have estimated that the design time for a portal could be reduced by two magnitude orders. The track and electrification infrastructure produced with SIA includes only normalized equipment and strongly reduces the catalog of systems installed, which also generates cost reductions in maintenance and stock for components. As a result, maintenance teams may achieve a shorter answer and repair time to failures and they can be more skilled in the system to maintain. One shortcoming of our work is the difficulty to get infrastructure information from the railway companies to feed databases. Usually they have their own catalog and inventory, and they are not very collaborative among companies. The European level initiative for railway interoperability could be an important step in this address. Achieving the collaboration of manufacturers of railway components could be another one. More effort will be devoted to both groups in a near future.