Stochastic and Linguistic Models of Planning Repair and Restoration Work on the Street-Road Network

The problems of developing network schedules for maintenance and rehabilitation work on the street-road network (SRN) are analyzed using stochastic and linguistic models, which are designed to describe activities with variable volumes and sequences, or with fuzzy input data. The input data required to produce a network schedule is described, and it is shown that these input data are probabilistic in nature. A probabilistic and fuzzy formalization of the lead time is proposed. For the stochastic variant of the network schedule parameterization, the corresponding probabilistic characteristics are calculated, taking into account the selection of distributions of each stage. For the linguistic parameterization of the model, it is proposed to calculate the sum and maximum of two or more fuzzy variables. An affiliation function type is defined and a procedure for constructing fuzzy stage realization time is proposed. The graphs of membership functions of volumes and intensities of works are given, and the task of searching for optimal estimates of the beta-distribution parameters for the best approximation of the membership functions is set. The procedure for computing the linguistic maximum that implements the generalization principle is described. The problem of critical path search is set and solved taking into account fuzzy character of model parameters which include fuzzy time of stages realization and fuzzy relations of stages precedence. An original approach to the realization of the modeling scheme is described, which is based on the principle of generalization followed by an extension of operations to find the time of stages and the volume of resources involved. The time and resource volume membership functions for the stages of the network schedule are derived.