The Design of Real-Time Expert Systems for Autonomous Vehicles

该提案旨在研究自主水下航行器（AUV）系统中基于知识的系统设计中的关键问题。它定义了理解与在实时系统中利用基于知识的系统相关的一些问题的努力。本研究的目的是为进一步开发基于实时知识的系统提供一个框架。该框架将通过开发适用于时间约束符号系统的操作系统和系统架构来实现。这项工作的主要成果将是一个新的符号处理操作系统，它将作为智能时间约束系统当前和未来发展的基础。提出的研究解决了导致有能力的实时系统的潜在问题。基于知识的技术将通过首先从操作系统和系统体系结构的优势来解决设计问题来推进。所提出的工作将利用当前的实时操作系统技术。一旦选择了架构和应用程序，它将在MSEL集成符号数字系统上实现，该系统位于用于VME总线的Ironics IV 3204 32位CPU板上。测试阶段将包括关键实时系统参数的确定和测量以及基准测试。在这一点上，除了简单的周期速率确定之外，没有明确的性能度量。由于这个原因，包含了一个任务来确定评估时间受限的基于知识的系统的有用度量。在这项工作的最后阶段，我们将评估结果，以确定性能限制并提出对系统的改进建议。

This proposal is for a research program that investigates key issues in the design of knowledge based systems for use in Autonomous Undersea Vehicle (AUV) systems. It defines an effort to understand some of the problems associated with utilizing knowledge based systems in real time systems. The purpose of this research is to implement a framework for further development of real time knowledge based systems. this framework will be realized with the development of an operating system and system architecture applicable to time constrained symbolic systems. The main product of this effort will be a new symbolic processing operating system which will be used as the foundation for current and future development in intelligent time constrained systems. The proposed research addressed the underlying issues which lead to capable real time systems. Knowledge based techniques will be advanced by first addressing the design from the vantage of operating systems and system architecture. The proposed work will take advantage of current real time operating system technology. Once the architecture a and application is chosen it will be implemented on the MSEL integrated symbolic numeric system residing on the Ironics IV 3204 32 bit CPU board for VME bus. The testing phase will include the determination and measurement of key real time system parameters as well as benchmarks. At this point there are no clear measures of performance other than simplistic cycle rate determinations. For this reason a task is included to determine useful metrics for evaluating time constrained knowledge-based systems. In the final phase of this work we will evaluate the results to determine performance limitations and propose enhancements to the system.