Застосування нечіткого алгоритму керування рухом на регульованому перехресті

Abstract

У роботі описано нечіткий алгоритм керування рухом на регульованому перехресті, керуючим рішенням якого є встановлення тривалості дозвільного сигналу в момент його ввімкнення. Встановлено, що застосування цього алгоритму є доцільним, оскільки дає змогу покращити якість та ефективність роботи перехрестя. Traffic lights are basic means of traffic control in cities. It is necessary to adjust function parameters of traffic lights depending on vehicle movement for efficient traffic control. So, improvement of existing or development of new algorithms for traffic control at signalized intersections is an urgent target. Control algorithm is a set of single-valued rules for processing data about traffic flow parameters at intersection, resulting in one or another control action [1]. Recently the algorithms with fuzzy logic have been applied in traffic control systems [2, 3]. The authors have improved fuzzy algorithm for traffic control, features of which are given in [4]. The determination of green light duration is the control action in this algorithm. Input linguistic variables are volume of arrival flow and queue length at some approach. Fuzzy control algorithm is implemented in MATLAB using Fuzzy Logic Toolbox. The model of signalized intersection Levytskoho str. – Tershakovtsiv str. – Dorosha str. (Lviv), developed in MATLAB, was chosen for research of intersection functioning with fuzzy control algorithm. Description of this model is given in [5]. Functioning of control system in which fuzzy algorithm is used (fuzzy control system) has been compared with functioning of fixed-time control systems: existing and calculated for given arrival flow volume. It was defined that the usage of fixed-time control system, calculated for given traffic conditions, will be the best, if arrival flow volume is invariable. But if the volume has an abrupt increase (even short-term) the usage of fixed-time control will not be reasonable as long as the value of the average and maximum queues has a sharp growth. However, fuzzy system adapts to traffic conditions at intersection well, and with a slight growth of queues at one approach smaller queues can be reached at other approaches.