【作者】 杨方燕；

【导师】 郭钢；

【作者基本信息】 重庆大学 ， 机械工程， 2018， 博士

【摘要】 对于汽车普通用户,特别是驾驶员来说,驾驶舱内的人机交互系统(如车载信息系统)是执行驾驶主任务与次任务的主要交互对象,对驾驶安全有重要影响。现有研究成果无法对面向安全驾驶的汽车人机交互进行全面测试评价,并指导产品创新设计。为此,本文提出一种基于用户(驾驶员)脑认知表征的面向安全驾驶的汽车HMI(人机交互界面)感知-人机交互愉悦测试评价及愉悦引导方法,力求为产品创新提供参考。本文依托于国家自然科学基金与“十二五”科技支撑计划相关课题,以乘用车驾驶舱车载信息系统HMI为对象,探索人-车-路系统中驾驶次任务驱动下,面向安全驾驶的驾驶员多模态人机交互环节中的脑认知机理和规律,对HMI感知-人机交互愉悦度的测试评价和愉悦引导方法与技术进行深入研究。主要研究内容如下:(1)面向安全驾驶的汽车多模态感知-人机交互脑认知模型研究。面向安全驾驶,对汽车多模态人机交互可感知特征向量进行提炼与标识;研究面向安全驾驶的汽车HMI脑认知表征愉悦建模技术,并对其影响因子进行分析;建立面向安全驾驶的汽车多模态感知-人机交互脑认知模型。(2)面向安全驾驶的汽车HMI愉悦测试评价方法与关键技术研究。建立汽车多模态人机交互下安全驾驶愉悦评价模型;研究汽车多模态人机交互下安全驾驶愉悦度测试方法与数据处理技术,建立多元数据特征量提取、统计、去噪、基于安全的GA-BP神经网络愉悦评价等核心算法,建立汽车多模态人机交互下安全驾驶愉悦度计算方法和量化评价准则。(3)面向安全驾驶的汽车多模态感知-人机交互愉悦引导方法研究。研究安全驾驶低愉悦度与汽车HMI可感知特征映射关联关系;建立汽车HMI可感知特征向量安全驾驶愉悦引导策略。(4)工程实例验证。在本文研究理论方法及关键技术的支撑下,以长安逸动EV300电动汽车车载信息系统HMI为例进行面向安全驾驶的汽车多模态人机交互愉悦测试评价与引导方法的应用,验证了本文方法的有效性。更多还原

【Abstract】 For ordinary users of automobiles,especially for drivers,the human-machine interaction system in the cockpit(e.g.vehicle information system),the main interactive object between the driver’s primary and secondary tasks,has an important impact on driving safety.Existing research results can’t meet the needs of the comprehensive testing and evaluation on in-vehicle human-machine interaction for safe driving,and help manufacture the products with innovative designs.For the reasons above,this Ph.D.thesis proposes a method of in-vehicle HMI perception-human-machine interaction pleasure testing and pleasure guidance for safety driving based on user’s(driver’s)brain cognitive representation,in order to provide a reference for product innovation.The thesis,relying on the scientific and technologic supports from the related subjects like the National Natural Science Foundation and the "Twelfth Five-Year Plan",taking the HMI in the cockpit as the research object,explores the brain cognitive mechanism and rules of driver-oriented multi-modal human-machine interaction(HMI)in driver-vehicle-road system driven by the secondary tasks,and the methods and techniques of testing and evaluating HMI perception-human-machine interaction pleasure and pleasure guidance.The main research content of the thesis is as follows:(1)Research on the brain cognition model with in-vehicle multi-modal perception-human-machine interaction for safe driving.The perceptive feature vectors of in-vehicle multi-modal human-machine interaction are extracted and identified for safe driving.The thesis investigates the pleasure modeling technology of HMI brain cognitive representation for safe driving and analyzes its influencing factors.An in-vehicle multi-modal perception-human-machine interaction and brain cognition model for safe driving is established.(2)Research on key technologies of pleasure testing and evaluation with in-vehicle HMI for safe driving.A pleasure evaluation model for safe driving under in-vehicle multi-modal human-machine interaction is constructed.The pleasure testing methods and data processing technologies by using in-vehicle HMI of multi-modal human-machine interaction are studied.The core algorithms such as feature extraction,statistics,denoising and pleasure evaluation based on GA-BP neural network are established.A method of calculating the pleasure degree of safe driving and a quantitative evaluation criterion for in-vehicle multi-modal human-machine interaction are established.(3)Research on pleasure guidance method of in-vehicle multi-modal human-machine interaction for safe driving.The mapping association relationship between low driving pleasure and perceived feature of HMI is studied.A pleasure guidance strategy of in-vehicle HMI sensible feature vectors for safe driving is established.(4)A case study.With the support of the theoretical method and the key technology of the thesis,the thesis takes the HMI of the on-board information system in EV300 as an example to carry out the evaluation and guidance of the in-vehicle multi-modal human-machine interaction pleasure testing for safe driving.The application of the guidance method verifies the effectiveness of the method.更多还原