中文摘要：

从驾驶员辅助系统与驾驶员的协作和协调角度,研究智能驾驶员辅助系统辅助级别判定的静态和动态判定方法、模型结构。提出了基于任务分解、功能划分的辅助级别判定原则和方法,建立了基于概率统计的效费评估模型,通过人机效费差值确定辅助级别。通过数学模型的模块化,在辅助系统人机接口中集成效费评估模型,实现了辅助级别的动态判定,为人机功能变自主权限管理提供依据。通过仿真试验验证了量化模型的工程合理性和可实现性。

英文摘要：

The aiding-level judging methods and architectures of intelligent assistant system were investigated from the perspective of system and human coordination and cooperation. The judging principles and methods based on task and function decomposition were proposed. Utility assessment modeling based on probability theory was realized. Aiding-level was decided by the difference of benefit-cost value between system and human. The mathematics model was integrated into the human-machine interface of assistant system for realizing the dynamic aiding-level judgment. The method was the fundament of function allocation based on variable autonomy level. The effectiveness and feasibility of quantitative model applied in engineering was demonstrated by the simulation.