【作者】 李绍松；

【导师】 宗长富；

【作者基本信息】 吉林大学 ， 车辆工程， 2013， 博士

【摘要】 汽车电动助力转向（Electric Power Steering，EPS）系统在提供转向助力、减轻驾驶员操纵负担的同时，也能够提高汽车转向性能和驾驶舒适性，进而提高汽车的主动安全性。目前，EPS系统在国内外已经成为很多车型的标配，但在EPS系统开发过程中，EPS控制参数的确定主要依赖调试人员的实车主观感觉，同时转向性能与EPS控制参数之间存在多参数和多性能的相互影响，在一个工况匹配好的控制参数，在另一个工况可能不适用，导致EPS控制参数实车调试和标定周期较长且效果不理想。因此，理清转向性能与EPS控制参数的影响关系，获得满足转向性能要求的EPS控制参数取值，为EPS控制参数实车调试和标定提供理论指导，能够有效加快EPS系统匹配进度，缩短EPS系统开发周期。结合国家高技术研究发展计划（863计划）课题―电动汽车底盘及整车控制关键技术‖（课题编号：2012AA110904）和吉林省重大科技专项―奔腾混合动力轿车电动助力转向系统匹配与开发‖（编号：20106001），本文搭建EPS系统与车辆模型联合仿真平台，通过模型仿真手段进行EPS控制参数与转向性能的灵敏度分析，理清转向性能与EPS控制参数间的敏感性关系，并在灵敏度分析研究结果基础上优化EPS控制参数，确定获得理想转向性能时的EPS控制参数。同时，考虑到EPS可以主动输出力矩，在EPS传统控制功能基础上，充分利用EPS系统现有传感器信号，开发EPS附加控制功能，在不增加EPS系统硬件成本情况下提高汽车转向性能。论文的具体研究内容如下：（1）搭建EPS系统与车辆模型联合仿真平台通过模型仿真手段从提高汽车转向性能角度对EPS系统进行研究，实现以车辆性能预测为基础开发EPS系统，能够有效节约EPS系统开发费用，缩短EPS系统研发周期。车辆模型是仿真平台的重要组成部分，为了实现对车辆系统的高精度描述，车辆模型选用CarSim商用软件；同时，为了方便对EPS系统的研究，基于Matlab/Simulink建立EPS系统模型，搭建EPS系统与车辆模型联合仿真平台。联合仿真平台的参数来源是保证模型精度的重要环节，为此基于K&C试验台对某车型车辆参数和EPS总成特性进行测试，并对EPS总成特性的测试方法进行详细描述。（2）基于转向性能的EPS控制参数灵敏度分析作为电子技术与转向系统相结合的产物，EPS系统在提高汽车转向性能方面有很大潜力。但在EPS系统开发过程中，对转向性能与EPS控制参数的影响关系认识不十分明确，EPS控制参数的确定主要依赖调试人员的主观感觉，导致EPS系统开发周期较长。因此，理清转向性能与EPS控制参数间的敏感性关系，掌握转向性能评价指标与EPS控制参数的影响关系，能够为EPS控制参数实车调试和标定提供理论依据和指导。本文对EPS控制参数与转向性能灵敏度分析时，确定了考察的转向性能及具体评价指标，给出了灵敏度分析的EPS控制参数及变化水平，最后基于Isight软件确定出EPS控制参数与转向性能的灵敏度关系。（3）基于转向性能的EPS控制参数优化在EPS控制参数与转向性能灵敏度分析研究结果基础上，筛选确定对转向性能敏感性较高的EPS控制参数作为优化对象，基于转向性能优化确定获得理想转向性能时的EPS控制参数，对加快EPS系统匹配进度，指导EPS系统开发有重要意义。基于转向性能优化EPS控制参数时，确定了汽车转向性能的优化目标，并在EPS控制参数与转向性能灵敏度分析研究结果基础上，确定优化的控制参数及约束条件。同时，针对转向性能与EPS控制参数间存在的耦合现象，建立转向性能的综合评价指标进行多目标优化，确定出满足各转向性能要求的EPS控制参数取值。（4）开发EPS附加控制功能充分利用EPS系统现有传感器信号，在不增加系统硬件成本基础上，开发EPS附加控制功能，实现以附加软件形式提高汽车转向性能，进而掌握EPS附加控制功能的核心技术，缩小国内EPS产品与国外同类产品的差距。针对无转角传感器情况下EPS汽车回正控制实现困难问题，提出无转角传感器下的EPS主动回正控制方法，基于卡尔曼滤波估计轮胎回正力矩，以此为基础确定回正控制补偿电流，改善汽车回正性能；同时，提出EPS衰减转向盘冲击力矩控制方法，采用高通滤波处理方法区分有用路感信息和路面冲击力矩，并在估计的路面干扰力矩基础上，确定路面冲击补偿控制电流，衰减路面冲击产生的转向盘冲击力矩，提高EPS汽车抗路面冲击性能，进而提高驾驶的舒适性。（5）搭建永磁同步电机EPS试验平台，对优化后的汽车转向性能和EPS附加控制功能进行实车试验验证采用矢量控制方法进行永磁同步电机的底层驱动控制，开发永磁同步电机EPS原型控制器，并针对国内某品牌车型进行EPS实车改装，搭建永磁同步电机EPS实车试验平台。以优化确定的EPS控制参数为基础，结合主观评价人员的主观评价结果，调整确定EPS原型控制器中的控制参数，并进行EPS转向性能客观评价试验，验证优化后EPS样车的转向性能。同时，针对开发的EPS附加控制功能设定试验工况，验证无转角传感器下的EPS主动回正控制和EPS衰减转向盘冲击力矩控制方法的实车控制效果。本文取得的创新性成果如下：（1）针对EPS系统开发过程中，对EPS控制参数与转向性能的关系不明确，控制参数的确定主要依赖调试人员的实车主观感觉，调试周期较长的现象，基于转向性能客观评价指标对EPS控制参数进行灵敏度分析及优化。确定EPS控制参数与转向性能评价指标的影响关系，在此基础上筛选确定对转向性能敏感性较高的EPS控制参数进行优化，确定满足转向性能要求的EPS控制参数取值，指导EPS控制参数的实车调试和标定，提高实车调试和标定效率。（2）助力特性是EPS基本助力控制实现的关键，体现的是EPS助力大小和方向随驾驶员手力和车速的变化规律。多点折线拟合曲线助力特性在折线型助力特性的基础上，通过增加节点数量，能够达到曲线型助力特性的效果，具有很好的实际应用价值。针对多点折线拟合曲线助力特性多个助力调节特征点设计难的问题，对多点折线拟合助力特性进行转矩区域划分，确定EPS助力特性具体转矩区域与转向性能的灵敏度关系，指导EPS助力特性参数的设计。（3）充分利用EPS系统现有传感器信号，开发EPS附加功能控制算法，在不增加EPS系统硬件成本基础上，以附加软件形式提高汽车转向性能。针对无转角传感器的EPS汽车回正难解决问题，在准确估计轮胎回正力矩基础上，进行无转角传感器下的EPS主动回正控制，改善汽车回正性能；同时，开发EPS衰减转向盘冲击力矩控制方法，提高EPS汽车的抗路面冲击性能，进而提高驾驶的舒适性。更多还原

【Abstract】 Electric power steering system not only can provide steering power and lighten theburden of the driver’s manipulation, but also can improve vehicle steering performance anddriving comfort, thus improving vehicle active safety. Currently, EPS system at home andabroad has been standard configuration on many cars. However, in the process of EPSsystem development, the determination of EPS control parameters mainly depends ondebugging personnel’s real vehicle subjective feeling. Meanwhile, the influence of multipleparameters and performance may exist between steering performance and EPS controlparameters. And matched control parameters in one working condition may be not applicablein another one, which lead to a long period of real vehicle debugging and calibration for EPScontrol parameters, and the effect may not be ideal either. Therefore, clarifying influencebetween steering performance and EPS control parameters, getting EPS control parametervalues meeting the performance requirements can provide theoretical guidance for realvehicle debugging and calibration of EPS control parameters, effectively accelerate thematching progress and reduce the development cycle of EPS system.Combined with the national high technology research and development project (863Program)―Research on Key Technologies of Chassis and Vehicle Control for Electricvehicle‖(Project number：2012AA110904) and major projects in Jilin Province―Researchon Key Technologies of selection and Development for Electric Power Steering System Base on Pentium Hybrid Car‖（Project number：：51105165）, the co-simulation platform of theEPS system and vehicle model is built in this paper and the sensitivity analysis of EPScontrol parameters and steering performance by means of model simulation is discussed.Then the sensitivity relationship between EPS control parameters and steering performanceis clarified and the EPS control parameters are choosed to be optimized based on theresearch results of the sensitivity analysis. Finally, the EPS control parameters are confirmedunder ideal steering performance. Meanwhile, considering that EPS can actively exporttorque, additional control functions are developed with full use of signals of EPS systemexisting sensor on the basis of traditional control function, which can effectively improvesteering performance without any increasement in the hardware cost.The specific contents of this dissertation are as follows:(1) A co-simulation platform of EPS system and vehicle model is builtThrough the aspect of promoting vehicle steering performance, research on the EPSsystem is conducted by means of model simulation. So EPS system can be developed basedon the prediction of vehicle steering performance, which can effectively save thedevelopment cost and decrease the development period of the EPS system. Vehicle model isan important part of the co-simulation platform, and commercial software named Carsim isused to accurately describe the vehicle system. Meanwhile, in order to make research on EPSsystem convenient, the EPS system is constructed based on Matlab/Simulink. Finally, theco-simulation platform of the EPS system and vehicle model is built. Considering thatparameters source of the co-simulation platform is also an important part to ensure the modelaccuracy, the parameters of vehicle and EPS assembly characteristics for one certain car aretested based on K&C test bench, and the testing method is also described in this paper.(2) Sensitivity analysis of EPS control parameters based on steering performanceAs the product of electronic technology combined with steering system, EPS systemhas a great potential in improving steering performance. However, during the developmentprocess of EPS system, the influence relationship between EPS control parameters andsteering performance is not clearly. And the determination of EPS control parameters are mainly depended on debugging personnel’s real vehicle subjective feeling, which increasethe development periodof EPS system. Therefore, clarify the sensitivity relationship betweensteering performance and the EPS control parameters, and grasp the effect of EPS controlparameters on the evaluation indexes of steering performance, which can provide theoreticalevidence and guidance for the regulation and calibration of EPS control parameters. Whensensitivity analysis of EPS control parameters and steering performance is conducted, firstly,steering performance and its specific evaluation indexes are determined. Secondly, EPScontrol parameters and their variation values are determined. Finally, the sensitivityrelationship between steering performance and EPS control parameters is clarified base onIsight software.(3) The optimization of EPS control parameters based on vehicle steering performanceBased on the sensitivity analysis results of EPS control parameters and steeringperformance, the EPS control parameters of higher sensitivity for steering performance arescreened and chosen as the optimization variables. Then EPS control parameters areoptimized to get the ideal steering performance, which can effectively accelerate thematching process of EPS system and has great significance on EPS system development.When optimization of EPS control parameters is conducted, Optimization object of EPSsteering performance is determined. And then control parameters for optimization areobtained based on the results of the sensitivity analysis. Meanwhile, considered the couplingeffects between steering performance and EPS control parameters, a comprehensiveevaluation index of steering performance is established for multi-objective optimization toget EPS control parameters which meet all the requirements of steering performance.(4) The development of EPS additional control functionsBy making full use of existing sensor signals of EPS system, EPS additional controlfunctions are developed without any increasement of the EPS system hardware cost toimprove steering performance by additional software, which is of great significance tomaster core technology of EPS additional control functions and can greatly narrow thetechnological gap between domestic EPS products with similar foreign products. For the difficult issue of return control for EPS vehicles without angle sensor, an EPS active returncontrol method without angle sensor is proposed. Tire aligning torque is estimated based onKalman Filter and compensation current for return control is determined to improveautomotive returnability. Meanwhile, a control algorithm for EPS vehicles to reduce theimpact torque on steering wheel is proposed by applying high-pass filtering processing todistinguish useful road sense information and road impact torque. Based on the estimateddisturbance torque from road surface, impact compensation control current is determined toimprove the road impact resistance of EPS vehicles so as to improve the driving comfort.(5) EPS real vehicle test platform is built to verify the optimized steering performanceand the developed additional control algorithm of EPS systemThe underlying drive control of permanent magnet synchronous motor is developed,and EPS prototype controller is also developed. Meanwhile, a domestic brand car ismodified and the real vehicle test platform is finally built. Combined with the subjectevaluation results of appraisers, the control parameters of EPS controller are determinedbased on control parameters determined by optimization. Then objective evaluation ofvehicle steering performance is tested to verify the steering performance after optimizationof EPS sample car. As for the developed additional control algorithm of EPS, test conditionsare set and the real vehicle tests are conducted to verify the effective of EPS additionalcontrol algorithm.So in conclusion, several innovative aspects proposed by this dissertation are asfollowed：(1) In EPS system development process, the relationship between EPS controlparameters and steering performance is not clear, and the determination of EPS controlparameters mainly depends on debugging personnel’s subjective feeling, which makes thedebug cycle of EPS control parameters longer. As for this phenomenon, the sensitivityanalysis and optimization for EPS control parameters is reasearched based on objectiveevaluation indexes of vehicle steering performance. So the relationship between EPS controlparameters and the exact evaluation indexes of steering performance can be clearly determined. And on this basis, EPS control parameters with higher sensitivity for vehiclesteering performance are selected to be optimized, and finally the value of EPS controlparameters are determined for ideal steering performance. This research can be usefully forthe debug and calibration of EPS control parameters on real vehicle and can greatly improvethe efficiency of debug and calibration.(2) It is the boost curve that reflects the variation of the EPS assisting magnitude anddirection with the input of drivers and vehicle speed. On the basis of broken line type assistcharacteristic, the assist characteristic of multi-line fitting curve can be realized throughincreasing the numbers of nodes, which has a good practical significance. However, it is verydifficult for the node design of the assist characteristic of multi-line fitting curve. As for thisproblem, the regional breakdown for the assist characteristic of multi-line fitting curve isdetermined, and the sensitivity relationship for the exact torque region of EPS assistcharacteristic and steering performance is finally determined, which can be supplied as theguidance for the design of EPS assist characteristic curve.(3) EPS Additional control algorithm is developed, with the full use of sensor signals ofEPS system. Therefore, vehicle steering performance can be greatly improved based on theform of software without increasing the system components. It is difficult for vehicle withoutangle sensor returning to center, which is also a hot research spot. So a return-to-centercontrol method without angle sensor is proposed to improve return performance. Meanwhile,steering wheel impact torque may be generated by road impact, which brought roughinfluence for driver. So a control method is proposed to reduce the impact torque on thesteering wheel caused by road impact, and to improve driving comfort.更多还原