复杂驾驶情境下基于信息融合的车载多维数字界面交互行为研究

As a key module of advanced driving auxiliary systems, the in-vehicle digital interface helps to improve driving safety and performance and has become a research hotspot in automotive engineering and human-machine interaction. However, the existing studies are limited by the monodimensionality of digital interface interaction, randomness of index extraction, and lag of evaluation mechanism. Thus, this project is targeted at the theories and applications concerning the interactive behaviors of in-vehicle multidimensional digital interface (IVMDI) under real-road scenarios. Firstly, the factor restraint relationships and designing rules of theoretical information coding about interactive behaviors of IVMDI will be explored. Based on information fusion methods, individual index sets will be extracted and multi-sensor index feature models will be constructed. Then based on fuzzy network analytical methods, an IVMDI driving safety and performance evaluation model will be built, and thereby the safety grades of IVMDI will be divided. Finally, the evaluation methods for multitask driving IVMDI-driver interactive behaviors under real-road scenarios will be clarified, and an IVMDI optimized design iteration strategy will be put forward. Thus, an IVMDI design - evaluation - redesign - application system based on information fusion will be established. The findings will provide novel theoretical methods for research on the decision-making and control of IVMDI interactive behaviors under complex driving scenarios and will raise the level of driving safety comprehensively.

车载数字界面作为先进驾驶辅助系统的重要模块，有助于提高驾驶绩效和安全，已经成为当前汽车工程和人机交互领域的研究热点。针对现有研究中存在数字界面交互维度单一、指标提取随机化、评价机制滞后等科学问题。本项目拟在复杂驾驶情境下，针对车载多维数字界面(In-Vehicle Multidimensional Digital Interface, IVMDI)交互行为的理论和应用问题，探究IVMDI交互行为理论信息编码中各因子约束关系和设计规律；基于信息融合方法，提取单项指标集，构建多传感器指标特征模型；基于模糊网络分析法，构建IVMDI安全性和驾驶绩效评价模型；基于评价结果，划分IVMDI安全等级。旨在掌握IVMDI多任务交互行为评价方法，提出IVMDI优化设计迭代策略，形成IVMDI智能交互体系，研究成果可为复杂驾驶情境下IVMDI交互行为决策与控制提供新的理论研究方法，全面提升驾驶安全水平。

车载数字界面作为先进驾驶辅助系统的重要模块，有助于提高驾驶绩效和安全，已经成为当前汽车工程和人机交互领域的研究热点。针对现有研究中存在数字界面交互维度单一、指标提取随机化、评价机制滞后等科学问题。.本项目拟在复杂驾驶情境下，针对车载多维数字界面(In-Vehicle Multidimensional Digital Interface, IVMDI)交互行为的理论和应用问题，探究IVMDI交互行为理论信息编码中各因子约束关系和设计规律；基于信息融合方法，提取单项指标集，构建多传感器指标特征模型；基于模糊网络分析法，构建IVMDI安全性和驾驶绩效评价模型；基于评价结果，划分IVMDI安全等级。旨在掌握IVMDI多任务交互行为评价方法，提出IVMDI优化设计迭代策略，形成IVMDI智能交互体系。.本项目通过IVMDI为载体，提出了一种IVMDI信息编码整合设计新思路；提供一种综合考虑绩效性和安全性的评价新方法；建立一种基于智能化驾驶交互新机制。有效的解决了IVMDI的人机交互界面整合设计问题；多源指标的信息融合度量统一方法问题；IVMDI智能迭代产品升级问题。.基于本项目研究成果，在用户体验设计测试、多通道交互模式、智能产品体验设计等方向，与吉利汽车、美的集团合作，进行成果转化，共同开发智能信息系统与智能产品，全面提升汽车驾驶安全性，智能家电产品体验性，有效的推进产品的智能化升级，为用户提供更为多元化选择的交互平台, 以满足用户多样化的交互行为，在人类日常行为中显得更为普遍且更为重要。