【作者】 潘梦鹞；

【导师】 刘桂雄；

【作者基本信息】 华南理工大学 ， 制造工程智能化检测及仪器， 2012， 博士

【摘要】 主动性的机动车运行安全状态监测新模式是机动车运行安全检测技术未来发展的必然趋势。论文“机动车运行安全监测新模式研究”系统地研究机动车运动姿态、动载荷、制动监测模型的机动车运行状态监测体系的理论基础、关键技术和实现方法，这对提高机动车安全运行的技术保障能力、减少交通事故，促进机械制造以及仪器科学与技术发展，具有重要的学术价值和实际意义。研究工作得到教育部新世纪优秀人才支持计划项目(No.NCET-08-0211)、广东省科技攻关项目(20090119)和国家自然科学基金项目(No.60472006)的资助。论文首先对国内外机动车运行安全状态监测的研究现状进行分析，指出目前机动车监测参数少，未能真实、动态地反映机动车安全运行状态监测的需要，信息得不到综合有效利用，从而确定论文研究任务。论文主要工作包括：㈠提出新监测模式下的指标体系及监测模式的整体构架。新监测模式指标体系包括监测机动车运动姿态(MAP，Motion Attitude Parameter)、动载荷(DLP，Dynamic LoadParameter)、制动性能(BPP，Braking Performance Parameter)三类关键参数，通过建立机动车MAP、DLP和BPP信息之间的关联性可获得一些传统其它系统无法测量的参数；借助新监测模式27项参数信息可以真实、动态地反映机动车安全运行状态，比以往更加全面、科学；提出机动车安全运行状态监测平台的构架、车载信息感知及控制终端总体结构，从信息感知、信息传送、应用开发三个功能层次对监测系统功能需求进行分析，讨论分析车轮MAP、车轮DLP和车轮BPP等关键测量模型的实现思路。㈡系统研究机动车运动姿态MAP监测技术。提出基于三维加速度传感正方体四顶点配置组成12加速度组件配置GFSIMU方案，推导出机动车车身MAP测量车身姿态角B, B, B、速度v Bx, v By,v Bz、加速度a B x, a By,a B z等数学模型；提出车轮MAP加速度计配置方案，推导车轮MAP测量车轮i (i=1~4)姿态角i, i, i、轮毂加速度aS Ti, aS Si,aS Ci、轮毂速度vS Ti, vS Si,vS Ci、和转动参量wi, wi数学模型；由车轮毂形状特点，设计圆形三角活动轮毂抓盘实验装置来安装车轮传感模快，既方便完成试验，又不影响试验原理及方法；选用具有良好信噪比的切向加速度传感器输出信号、非常适合作为车轮信号特征量，纠正以往倾向选用向心加速度传感器输出信号的做法。机动车运动姿态MAP监测方案具有实际加速度传感器少、算法简单，求解容易，测量精度高特点。㈢系统研究车轮动载荷Li、制动器制动力Fb i的监测新方案与监测机理。动载荷Li的监测方案基于机动车行驶过程中，车轮动载荷Li受轴荷转移、路面不平整、地面反力影响而发生变化，推导出车轮动载荷Li数学计算公式。制动器制动力Fb i的监测方案基于机动车行驶过程中，车轮BPP的车轮滑移率S i、制动器制动力Fb i及平衡参数都会发生变化，推导出车轮制动器制动力Fb i及其平衡的数学表达式。车轮动载荷与制动器制动力的监测新方案较以往其它方法，具有安装容易、算法简单、真实、动态地反映整体车轮动载荷、制动信息的特点。㈣研究基于神经网络的车身角速度姿态角算法。在模型不明确、方法比较复杂的情况下，用神经网络模型去代替这些原来基于机理所建的模型，从而达到提高建模准确性以及减少算法误差。提出基于典型的路况车身角速度神经网络分割方法，考虑隐藏层神经元数目、训练样本数目和训练函数方面因素，研究基于正交设计的网络结构优化方法；设计状态分类器辨识相应运动状态，使数据顺利进入相应子神经网络进行解算。㈤开展机动车运行安全试验系统、监测实验平台研发与实验。车轮智能传感节点由双轴加速度传感器、单轴加速度传感器组合测量，通过两个传感器布置于电路板两面使三个加速度敏感轴交于一点；研发机动车GPS地理位置参数监测装置，开发包括远程数据服务器、SQL数据库、监测平台远程管理软件在内的基于Web的监测平台远程管理系统；利用实验条件，制订出测试方法。实验结果表明，监测系统及平台能够较准确地测出所有的机动车安全运行状态监测参数，并可实现远程在线监控功能。㈥提出将基于轮载式智能传感的机动车运行安全状态监测平台技术的其它拓展应用方案。基于轮载式智能传感机动车安全状态快速检测系统，既可节约场地，又可现场检测，检测能力比传统固定检测线提高1倍以上；基于轮载式智能传感四轮定位参数测量系统，实现在现场对车轮四轮定位参数的快速、动态、准确、真实检测；基于轮载式智能传感动平衡参数测量系统，实现对车轮动平衡参数的快速、动态、准确检测。论文实验表明，实验条件下车身俯仰角θB平均误差0.0690o，车身侧倾ФB平均误差0.0351o，车身前向νBx速度平均误差1.3607m/s，车轮外倾γi平均误差0.0734°，轮毂切向νSti速度平均误差0.1480m/s，车轮角速度ωi平均误差0.5420rad/s，相对误差2.25%。本文提出新的监测模式是可行的，传感器配置方案是正确的，监测参数是全面的，能够测出所有的机动车安全运行27项状态监测参数，达到预想效果，具有重要推广价值。更多还原

【Abstract】 The active novel mode of vehicle operating safe state monitoring will be the inevitabletrend of vehicle driving safety measurement technology development in the future. The thesis“Study on the novel mode of vehicle operating safe state monitoring” systematically studiesfundamental theory&key technology&realize method of vehicle operating safe statemonitoring system, including vehicle operating attitude monitoring model, vehicle dynamicload monitoring model, vehicle braking performance monitoring model, which has importantacademic value and practical significance in improving the technical protection of vehiclesafe operation, reducing traffic accidents and promoting the development of automobileindustry and Instrument Science and Technology. This paper was supported by the Programfor New Century Excellent Talents in University of Ministry of Education(No.NCET-08-0211),Guangdong province Science and Technology Researching Project (No.20090119) and Natural science Foundation of china (No.60472006).This paper analysed the research status in vehicle operating safe state monitoring athome and abroad and suggested that the vehicle monitoring parameters are few, which cannotreflect the requirement of vehicle operating safe state monitoring truly and dynamically andget information be used comprehensively and effectively, then made the research scheme.Main works:⑴An index system in novel monitoring mode and the whole structure of the monitoringmode were put forward. The novel monitoring mode index system include monitoringvehicle’s three key parameters: motion attitude parameter(MAP), dynamic load parameter(DLP) and braking performance parameter(BPP).By establishing the connection among MAP,DLP and BPP information, some parameters that other traditional systems cannot measure canbe obtained; With the27-parameter information in novel monitoring mode, vehicle operatingsafe state can be reflected truly and dynamically, which is more comprehensive and scientificthan ever. Proposed the structure of motor vehicle operating safe state monitoring platformand the over-all structure of vehicular information perception and control terminal; analysedthe functional requirement of monitoring system from three function levels of the informationperception, information transmission and application development; and discussed the ideason how to realize key monitoring models such as wheel MAP, wheel DLP, wheel BPP andso on; to lay foundation for deep research into the model project(program).⑵The monitoring technology of motor vehicle MAP was studied systematically anddeeply. GFSIMU scheme were proposed based on the3d acceleration sensor cube to configurate12-acceleration components, the mathematical modes of body attitude angleB, B, B, velocity v Bx, v By,v Bz and acceleration a Bx, a By,a Bz were deduced and then tohave the motor vehicle body MAP inertia be measured. Wheel MAP accelerometer schemewere proposed, and mathematical models of the wheel i(i=1~4) attitude angle i, i, i, thewheel hub acceleration aS Ti, aS Si,aS Ci, the wheel hub velocity vS Ti, vS Si,vS Ci, and therotating parameter were deduced and then to have the wheel MAP inertia be measured.Basing on the characteristic of the wheel hub’s shape, a circular triangle mobile hub controlplate was designed to install wheel sensor module, which is convenient to perform the test andalso of no impact on the principle and method of experiment. The output signal of tangentialacceleration sensor with better SNR is proposed to be chosen as the suitable feature variableof wheel signal, which could correct past practice that choosing centripetal acceleration sensoroutput signal. MAP monitoring program has the characteristics of less actual accelerationsensor, simple algorithm, easy to solve and high accuracy.⑶The novel monitoring scheme and principle for dynamic loadLi and the brakingforceFb iwas studied systematically. The monitoring scheme of dynamic loadLi is based onthe fact that wheel dynamic loadLi varies with the axle load’s transfer, road’s smoothnessand the ground counter-force’s influence, then deduced the mathematical computationalformula of wheel dynamic loadLi. The monitoring scheme of braking forceFb iis based onthe variation of wheel BPP’s wheel slip rateS i, braking forceFb iand balance parametersduring driving process of the motor vehicle, then deduced the wheel braking forceFb iand itsbalanced mathematical expression. Compared to other methods, this one’s installation is easyand its arithmetic is simple, as well, it could reflect the wheel braking information real-time,truly and dynamically.⑷The algorithm of vehicle body angular velocity attitude Angle was studied basing onneural network. While the model is not clear, the method is more complex, the neural networkmodel was used to replace the original model based on mechanism, so as to improve theaccuracy and reduce the algorithm error. The neural network splitting method was proposedbasing on the typical road-condition wheel body angular velocity to consider the number ofneurons, training sample numbers and training function factors in hidden layer. The researchbased on orthogonal design network structure optimization methods and designed the statesorter which could recognize the corresponding motion state, so as to make corresponding data smoothly enter the sub-neural network to calculate.⑸Research, development and experiment of the vehicles driving safe testing systemand monitoring experimental platform were carried out. The wheel intelligence sensor nodescombined biaxial acceleration sensors and monaxial acceleration sensors to form multiplemeasurement instruments. By placing two measuring sensors in two sides of the circuit board,three acceleration sensitive axis were intersected to one point. Based on WEB to research anddevelop the vehicles’ GPS location parameters monitoring device, develop the monitoringplatform remote management system, including remote data server, SQL database, monitoringplatform remote management software. Take advantage of the experimental conditions tomake out test method. Through experiment verification, the monitoring system and platformcould accurately detect all the monitoring parameters of vehicles operating safe state, andrealize the remote online monitoring function.⑹Based on WEIS other development application solutions of the vehicle operating safestate monitoring platform technology was proposed. The safe state rapid detection system thatbased on the wheel load type intelligent sensor WEIS can save space and have the fielddetection, its testing ability was1times above than traditional fixed detection line. Thefour-wheel location parameter measurement system that based on the wheel load typeintelligent sensor WEIS could provide rapid, dynamic, accurate and true detection forfour-wheel location parameters on the site. Based on the wheel load type intelligent sensorWEIS, the dynamic balance parameter measuring system could provide rapid, dynamic andaccurate detection for wheel dynamic balance parameter.The experiments show that, the body longitudinal attitude θBaverage error is0.0690o,thebody side-inclination angle ФBaverage error is0.0351o,the body forward direction velocityνBxaverage error is1.3607m/s, the wheel extraversion γiaverage error is0.0734°,the wheeltangential velocity νStiaverage error is0.1480m/s,the wheel angular velocity ωiaverage erroris0.5420rad/s, relative error is2.25%under the experimental conditions.the novel monitoringmodel on this paper was feasible, the sensor configuration scheme was correct, and themonitoring parameters were comprehensive which could detect all27of state monitoringparameters to reach the expected effect, with important promotional value.更多还原