Railroad track technological construction process modeling and optimization using a genetic algorithm

Abstract

The article discusses theoretical and practical research in the field of evolutionary modeling application for solving construction works planning problems in the formation of railroad track technological construction process. The study is a part of the developed engineering and technical support subsystem for railway construction: engineering and intellectual support of railroad track technological construction process. This subsystem is based on the effective use of automated systems with artificial intelligence elements. Development and implementation of these automated systems are focused on achieving a single end result, such as a finished railroad track facility with suitable functionality within the established deadlines, planned working cost and labor costs, as well as meeting modern safety requirements throughout the entire operation period. Taking into account railway construction dynamic, the use of new materials and technologies, new trends were identified within the study scope in the field of technological processes operational development, and optimal work sequence determination in the railway facilities construction. In particular, the occurrence of deviations from the planned requirements in the construction work course requires a quick reconsideration of the decisions already made, due to the railway construction stochastic behavior. This is a forced measure to ensure the fulfillment of the planned targets. The existing methods allow to correct the construction work organization, however, the technology, and in particular, the technological process remains unchanged. The static nature of the technological process is largely dictated by the project documentation requirements and work safety. But modern advances in the information and intelligent technologies field, with the account of technological process structure variability, can give a technological process dynamic properties. The technological process's ability to adapt to changing work production conditions will provide flexibility in the railway facilities construction. To solve this problem, an automated evolutionary modeling mode and optimization of the railway facility construction technological process were used. The peculiarities of existing optimization methods, in terms of the dimension of the active task and taking into account several criteria, force us to turn to intelligent methods. The article describes a technological processes optimization method for the railway facilities construction using a genetic directed search algorithm in the space of solutions. In this case, several design constraints are taken into account: resource, technological, organizational, informational. For this, a computational and logical model was developed, which made it possible to assess the target function (fitness function), with the account of the dynamic distribution nature of the contractor's available resources for construction work. Based on the theoretical research results, the article presents the practical aspects of evolutionary modeling and optimization of the railway facilities construction technological process using a genetic algorithm on the example of the flooded railway roadbed embankment construction. The results presented in this article were obtained during the dissertation research made by the author.