Research of intelligent driving support system for welfare vehicle operation

诸如电动四轮车辆的福利设备的开发在进步，并且这些高效设备的普遍使用已经被积极地进行。实现护理人员现在使用计算机进行的适当支持的系统是必要的，以便非专家和残疾人可以安全地使用具有这种先进性能的设备。在开发该系统时，如何在尊重人的意图的同时，支持人的驾驶和操作，是一个需要解决的重要问题。为了解决这一问题，本研究建立了人的驾驶模型，并采用模糊控制技术设计了允许非熟练驾驶的智能驾驶辅助系统的硬件和软件系统。计算机被安装在四轮车上是为了人类能够真正上车的福利。开发基于上车人的简单指导和适当的汽车操作的考虑周围环境和当前位置的智能驾驶系统。2)系统的开发，假设人直接操作的情况，通过理解非熟练操作者的特征，建立支持者的知识，并通过语音进行适当的运行支持的传播。3）提出了过去依赖于人的运行知识的设置，应用了强化学习的方法，计算机提出了自动获取知识的方法。此外，还内置了专家知识，能够识别驾驶员的驾驶操作，并通过语音和图形显示向驾驶员发出适当的引导。改进了与用户的接口，并通过对真实的车辆的综合评价，验证了所开发的智能驾驶支持系统的有效性。

The developments of a welfare equipment such as electric four-wheeled vehicles advances, and the popular use of these efficient equipment have been done actively. The system that achieves the appropriate support that man who is nursing is doing now with the computer is necessary so that non-expert and the handicapped person may safely use the equipment with this advanced performance. When this system is developed, it is an important problem that how man's driving and operation are supported while especially esteeming his intention should solve. To solve this problem in this research, man's driving model was made to exist, and hardware and software system of the intelligent driving support that applied the fuzzy control technique that allowed an unskilled driving was made.First of all, the computer was installed in the four-wheeled vehicle for welfare that man was able to actually get on. 1) Development of intellectual driving system that considers surrounding circumstances and present location based on easy instruction of got on man and operates appropriate car. 2) Development of system that assumes case where man operates it directly, builds in knowledge of those who support it by understanding of non-skill operation person's characteristic, and does dissemination of appropriate running support by voice, 3) It proposes the setting of the running knowledge to rely on people in the past, the method of reinforced study is applied and the computer proposes the method to acquire knowledge automatically. These effectiveness was confirmed.In addition, expert's knowledge was built in, man's driving operation was recognized, and an appropriate guidance was notified to man by a voice and a graphic display. The interface with the user was improved, and the effectiveness of the intellectual driving support system that developed was confirmed by the comprehensive evaluation that used a real vehicle.