1-407 適応ニューラルネットワークを用いたスリップ制御とプラトーン安定性

概要

論文の詳細を見る
AHS (Automated Highway Systems) project is one of the ITS (Intelligent Highway Systems) discipline that has developed all over the world. The platoons control problem, the longitudinal vehicle tracking control, is important AHS project. The control schemes to the problem are "vehicle follow-up control" in which the preceding vehicle's position and velocity are the feedback information. The controlled system on each vehicle consists of an upper level, from the throttle or brake to acceleration, and a lower level, from acceleration to position. Traditional researches to the problem assume that vehicle parameters are known and time invariant, and do not consider the slip between the tires and the ground for simplicity in many cases. However, the parameters are nonlinear, unknown and time variant, and the slip does actually occur in emergency situations such as a sudden braking. Conventional method, which focuses solely on robustness, is inadequate. In this paper, we will present an ANN (adaptive neural network) control scheme, which considers variable parameters and slip. Employing ANN controller, with the upper level controller, we will mathematically prove that the presented methods guarantee the platoon stability in which spacing errors are not amplify as they propagate towards the tail of the string.
社団法人日本機械学会の論文
2002-11-14