作者: Xin Wang , M. Sugisaka
关键词:
摘要: The goal of the paper is to submit an idea from practical point view for control technologies. Our basic make a mobile vehicle (NMV) be capable evaluating results actions and modifying inputs controller itself in order drive optimal or near-optimal paths. NMV system consists charge-coupled device (CCD) camera, Von Neumann type computer neurocomputer, image processing units etc. hardware neurocomputer RN-2000 kernel part system. purpose solve problem how realize by backpropagation (BP) neural network learning online. strategy presented based on past patterns adjusting content driving new algorithm. Learning happens during procedure vehicle. This research shows possibility which BP inside can learn human knowledge online aid software. Some words about our researches are also given help understanding method.