
Тема выпуÑкной квалификационной работы: «Ðнализ и улучшение организации трафика Ñ Ð¿Ð¾Ð¼Ð¾Ñ‰ÑŒÑŽ иÑкуÑÑтвенных нейронных Ñетей». Ð’ данной работе раÑÑматриваетÑÑ Ð²Ð¾Ð¿Ñ€Ð¾Ñ ÑÐ¾Ð·Ð´Ð°Ð½Ð¸Ñ ÑиÑтемы умных Ñветофоров, иÑпользующих технологию нейронных Ñетей Ð´Ð»Ñ Ñ€ÐµÐ³ÑƒÐ»Ð¸Ñ€Ð¾Ð²Ð°Ð½Ð¸Ñ Ð¸ Ð¾Ð¿Ñ‚Ð¸Ð¼Ð¸Ð·Ð¸Ñ€Ð¾Ð²Ð°Ð½Ð¸Ñ Ð´Ð¾Ñ€Ð¾Ð¶Ð½Ð¾Ð³Ð¾ трафика. Цель работы: разработка нейронной Ñети Ð´Ð»Ñ ÑƒÐ¿Ñ€Ð°Ð²Ð»ÐµÐ½Ð¸Ñ Ñветофорами. Результаты работы: работа была поделена на два больших Ñтапа. Ðа первом, «локальном», Ñтапе была Ñоздана Ð½ÐµÐ¹Ñ€Ð¾Ð½Ð½Ð°Ñ Ñеть Ð´Ð»Ñ Ð¾Ð´Ð½Ð¾Ð³Ð¾ Ñветофора, контролирующего только один перекреÑток. ÐÐµÐ¹Ñ€Ð¾Ð½Ð½Ð°Ñ Ñеть обучилаÑÑŒ и результатом Ñтапа Ñтало Ñнижение времени проÑÑ‚Ð¾Ñ Ð½Ð° 37% по Ñравнению Ñо Ñтандартным Ñветофором Ñ Ñ„Ð¸ÐºÑированным временем Ñмены Ñигналов. Второй Ñтап заключалÑÑ Ð² Ñоздании полноценной транÑпортной Ñети. Ðа втором Ñтапе работы умные Ñветофоры (Ñозданные на первом Ñтапе) были объединены в единую Ñеть. Ð ÐµÐ°Ð»Ð¸Ð·Ð¾Ð²Ð°Ð½Ð½Ð°Ñ ÑиÑтема показала общее улучшение ÑоÑтоÑÐ½Ð¸Ñ Ñ‚Ñ€Ð°Ð½Ñпортной Ñети. При раÑÑмотрении 6-и чаÑового периода ÑреднÑÑ Ð·Ð°Ð´ÐµÑ€Ð¶ÐºÐ° каждого транÑпортного ÑредÑтва была Ñнижена на 26%, а ÑреднÑÑ ÑкороÑть выроÑла на 19%. Как итог, были доÑтигнуты многообещающие результаты, показывающие выÑокую ÑффективноÑть данного подхода к решению поÑтавленной задачи. ОблаÑть применениÑ: управление потоками дорожного транÑпорта в уÑловиÑÑ… городÑкой Ñреды.
Theme of this final qualifying work is "Analysis and improvement of traffic organization using artificial neural networks." This paper considers the issue of creating a system of smart traffic lights that use neural network technology to regulate and optimize road traffic. The aim of the work: development of a neural network for controlling traffic lights. The results: the work was divided into two large phases. At the first, “localâ€, phase a neural network was created for one traffic light (that controls only one crossroad). The neural network was trained and the result of the first phase was a 37% reduction in downtime (at a crossroad) compared to a standard traffic light with a fixed signal change time. At the second phase of work, smart traffic lights (those were created at the first stage) were combined into a network. The implemented system showed an overall improvement in the state of the transport network. Considered a 6-hour period, the average delay of a vehicle was reduced by 26%, and the average speed increased by 19%. In conclusion, promising results were achieved, showing the high efficiency of this approach for solving the problem. Scope of application: management of streams of road transport in the conditions of the urban environment. 
нейÑоÑеÑи, road traffic management, reinforcement learning, обÑÑение Ñ Ð¿Ð¾Ð´ÐºÑеплением, ÑвеÑоÑоÑÑ, traffic lights, ÑпÑавление доÑожнÑм движением, neural networks
нейÑоÑеÑи, road traffic management, reinforcement learning, обÑÑение Ñ Ð¿Ð¾Ð´ÐºÑеплением, ÑвеÑоÑоÑÑ, traffic lights, ÑпÑавление доÑожнÑм движением, neural networks
| selected citations These citations are derived from selected sources. This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically). | 0 | |
| popularity This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network. | Average | |
| influence This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically). | Average | |
| impulse This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network. | Average |
