Constructing a model for the automated operative planning of local operations at railroad technical stations

This paper has investigated the technology of forwarding local wagons at railroad technical stations and established need to improve it given extra downtime wagons. The main issue relates considerable combinatorial complexity tasks operational planning. Another problem is that as part conventional approach, planning a station operation considered separately. lack high-quality models for preparation initial data, in particular, data on duration technological operations, such as, example, shunting operations involving forwarding. To resolve these issues, new approach been proposed, under which operative station’s its subsystem are tackled simultaneously, based single model. this end, mathematical model vector combinatoric optimization built, uses criteria total operating costs wagon-hours spent when wagon flows, form separate objective functions. Within model, predictive was constructed fuzzy inference system. designed determine half-runs executing spotting/picking delivering enterprises’ goods sheds. provides accuracy level would suffice planning, contrast classical methods. A procedure devised optimizing employs modern genetic algorithm NSGA-III. implemented software makes possible build rational plan station, including graphic form, thereby reducing by 5 % maintenance 8 %. resulting effect could reduce turnover time freight car general network, speed up delivery goods, cost transportation