Adaptive Interface of Considering the Chaos Dynamics in Human Skill

在本计画中，我们以借由考虑人与环境之间的混沌动力学来提高介面的适应性为目标，将步行机器人、人工神经网路视觉系统与电动轮椅车作为一个整体系统来开发。主要研究结果如下：1.通过将混沌理论应用于步行和语音中的测量信号，我们清楚地了解了人体运动中的非线性吸引动力学。开发了具有中央模式发生器神经元的两足步行机器人.人工神经网络视觉智能信息处理系统的构建将脉冲耦合神经网络应用于初级视觉处理，开发了具有目标分割、视觉聚焦、目标跟踪等功能的仿真系统.考虑人机交互动力学的电动轮椅自适应接口的研制我们研制了电动轮椅车系统，用于测量和控制人的动力学行为。该系统配备22个传感器，测量数据同步处理，采用自适应运算。作为一个例子，人与汽车之间的不良互动，我们展示了自适应接口的方法，以防止驼背现象。

In this project, focusing the target to improve the adaptivity of interfacing by considering the chaos dynamics between human and environment, we developed the walking robot, the artificial neural network vision and the electric wheel chair car as a total system. The results are summarized in the followings ;1. Evaluation of the chaotic skill dynamics within the human walking and speech By application of chaos theory to the measured signal in walking and speech, we made clear of the nonlinear attracting dynamics within the human motions. We developed the 2-legged walking robot with the central pattern generator neurons.2. Constitution of the articial neural network vision with intelligent information processing Applied the pulse coupled neural networks to the primary visual processing, we developed the simulated system which have the capability to process the segmentation, foveation, target tracking of the objects.3. Development of the adaptive interfave of electric wheel chair of considering the interactive dynamics between human and machine We developed the electric wheel chair car system to measure and control the dynamical behaviors with human. The 22 sensors are equipped in this system and the measured data are processed synchronously to utilize the adaptive operation. As an example of the undesirable interaction between human and car, we showed the adaptive interfacing method to prevent the hunching phenomenon.