Further studies dynamic neural networks and fuzzy logic with applications to neuro-control and neuro-vision systems in robotics

动态神经网络和模糊逻辑是近年来发展起来的两种理论工具，在复杂动态系统控制、视觉系统、机器人、医学图像处理、工业系统决策等领域有着广泛的应用。在过去的二十年里，我们在智能系统研究实验室的团队在这两个重要领域发展了广泛的理论基础，融合了这两个用于开发模糊神经系统的数学工具，并将这些理论工具应用于复杂动态系统的问题。在接下来的五年里，我们希望将这些理论工具扩展到动态神经网络和模糊神经系统领域，并将它们应用于一些工业、医学图像处理和空间问题。动态神经结构的进一步理论发展：将在开发新的动态神经结构、研究其联想记忆能力和稳定性特性方面开展理论工作。这种新颖的动态神经网络设计将引导我们研究它们的动态特性，如稳定性条件和平衡能力。具有新的动态神经结构并具有学习和适应能力的神经控制器将被开发出来。将在设计新的神经结构方面开展进一步的工作，以便将其应用于感受野、联想记忆、视觉记忆和其他视觉功能。同过去一样，学院的几名研究生、博士后研究员、客座教授/科学家和几名教职员工将在这些拟议研究计划的活动中发挥重要作用。加拿大这个国家需要一些在神经网络、模糊逻辑和机器人系统领域具有高水平研究经验的合格人才，像过去一样，这项拟议的研究承诺培养这样的合格人才。

Dynamic neural networks and fuzzy logic are the two recently developing theoretical tools which have found a wide range of applications in the areas such as control of complex dynamic systems, vision systems, robotics, medical image processing, and decision making problems in industrial systems. Over the past two decades, our group at the Intelligent Systems Research Laboratory has developed an extensive theoretical basis in these two important areas, have merged these two mathematical tools for developing the fuzzy neural systems and have applied these theoretical tools to problems of complex dynamic systems. Over the next five years, we wish to extend these theoretical tools in the field of dynamic neural networks and fuzzy neural systems, and apply them to some industrial, medical image processing and space problems. Further theoretical development of dynamical neural structures: Theoretical work will be carried out in the development of new dynamical neural architectures, study of their associative memory capabilities and stability characteristics. The novel design of dynamic neural networks will lead us to study their dynamic properties such as stability conditions and equilibrium capacity. The neuro-controllers with new dynamic neural structures and with learning and adaptive capabilities will be developed. Further work will be conducted on the design of new neural architectures for their applications to receptive fields, associative memory, visual memory and other visual functions. As during the past, several graduate students, postdoctoral fellows, visiting professors/scientists and several faculty members of the College will play an important role in the activities of these proposed research plans. This country Canada is in need of some qualified persons with high-level of research experiences in the field of neural networks, fuzzy logic and robotic systems, and as during the past, this proposed research has promise the produce such qualified personnel.