Traffic signal networks simulator with learning emotional algorithm
説明
Proposes traffic signal networks, which optimize the traffic flow by estimation of the driver's mind. The proposed simulator applies the emotional algorithm to determine the performance of each traffic signal individually and to avoid the traffic jam around it. The emotional algorithm estimates the emotional value based on the emotional functions, which are defined on the emotional space. The emotional space consists of four factors, "happy" "relief," "afraid" and "angry," and provides the emotional states of the system by the value of it. The emotional function evaluates and imitates the state of driver's mind, and aims at not inflicting psychological stress on the driver. This report proposes the principal architecture of the emotional algorithm and some simulators to use it to decide its performance of the traffic signal system. The numerical simulations clear the effectiveness of the proposed emotional algorithm and the applied system. The learning simulator, of which parameters dynamically change by evaluating the congestion of a road, especially shows the excellent performance. Date environment analysis is a good learning method to optimize the parameters in the emotional algorithm by simulation results.
収録刊行物
-
- Proceedings. 2000 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2000) (Cat. No.00CH37113)
-
Proceedings. 2000 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2000) (Cat. No.00CH37113) 3 2274-2279, 2002-11-08
IEEE