Performance optimization of electric power steering based on multi-objective genetic algorithm

The vehicle model of the recirculating ball-type electric power steering (EPS) system for the pure electric bus was built. According to the features of constrained optimization for multi-variable function, a multi-objective genetic algorithm (GA) was designed. Based on the model of system, the quantitative formula of the road feel, sensitivity, and operation stability of the steering were induced. Considering the road feel and sensitivity of steering as optimization objectives, and the operation stability of steering as constraint, the multi-objective GA was proposed and the system parameters were optimized. The simulation results show that the system optimized by multi-objective genetic algorithm has better road feel, steering sensibility and steering stability. The energy of steering road feel after optimization is 1.44 times larger than the one before optimization, and the energy of portability after optimization is 0.4 times larger than the one before optimization. The ground test was conducted in order to verify the feasibility of simulation results, and it is shown that the pure electric bus equipped with the recirculating ball-type EPS system can provide better road feel and better steering portability for the drivers, thus the optimization methods can provide a theoretical basis for the design and optimization of the recirculating ball-type EPS system.

HUMAN-SIMULATING VEHICLE STEERING CONTROL ALGORITHM

A new vehicle steering control algorithm is presented. Unlike the traditional methods do, the algorithm uses a sigmoid function to describe the principle of the human driver＇s steering strategy. Based on this function, a human simulating vehicle steering model, human-simulating steering control（HS） algorithm is designed. In order to improve the adaptability to different environments, a parameter adaptive adjustment algorithm is presented. This algorithm can online modify the value of the key parameters of the HS real time. HS controller is used on a vehicle equipped with computer vision system and computer controlled steering actuator system, the result from the automatic vehicle steering experiment shows that the HS algorithm gives good performance at different speed, even at the maximum speed of 172 km/h.

PARAMETER OPTIMIZATION OF ELECTRIC POWER STEERING INTEGRATED WITH ACTIVE FRONT STEERING FUNCTION

The dynanaic model of a novel electric power steering（EPS） system integrated with active front steer- ing function and the three-freedom steering model are built. Based on these models, the concepts and the quanti- tative expressions of road feel, sensitivity, and operation stability of the steering are introduced. Then, according to constrained optimization features of multi-variable function, a genetic algorithm is designed. Making the road feel of the steering as optimization objective, and operation stability and sensitivity of the steering as constraints, the system parameters are optimized by the genetic and the coordinate rotation algorithms. Simulation results show that the optimization of the novel EPS system by the genetic algorithm can effectively improve the road feel, thus providing a theoretical basis for the design and optimization of the novel EPS system.

Integration optimization of novel electric power steering system based on quality engineering theory

The dynamic model of a novel electric power steering （EPS） system integrated with active front steering function （the novel EPS system） is built. The concepts and quantitative expressions of the steering road feel, steering sensibility, and steering operation stability are introduced. Based on quality engineering theory, the optimization algorithm is proposed by integrating the Monte Carlo descriptive sampling, elitist non-dominated sorting genetic algorithm （NSGA-II） and 6-sigma design method. With the steering road feel and the steering portability as optimization targets, the system parameters are optimized by the proposed optimization algorithm. The simulation results show that the system optimized based on quality engineering theory can improve the steering road feel, guarantee steering stability and steering portability and thus provide a theoretical basis for the design and optimization of the novel electric power steering system.

Multi-objective optimization of active steering system with force and displacement coupled control

A novel active steering system with force and displacement coupled control(the novel AFS system) was introduced,which has functions of both the active steering and electric power steering.Based on the model of the novel AFS system and the vehicle three-degree of freedom system,the concept and quantitative formulas of the novel AFS system steering performance were proposed.The steering road feel and steering portability were set as the optimizing targets with the steering stability and steering portability as the constraint conditions.According to the features of constrained optimization of multi-variable function,a multi-variable genetic algorithm for the system parameter optimization was designed.The simulation results show that based on parametric optimization of the multi-objective genetic algorithm,the novel AFS system can improve the steering road feel,steering portability and steering stability,thus the optimization method can provide a theoretical basis for the design and optimization of the novel AFS system.

Multi-objective Optimization of Differential Steering System of Electric Vehicle with Motorized Wheels

A differential steering system is presented for electric vehicle with motorized wheels and a dynamic model of three-freedom car is built.Based on these models,the quantitative expressions of the road feel,sensitivity,and operation stability of the steering are derived.Then,according to the features of multi-constrained optimization of multi-objective function,a multi-island genetic algorithm(MIGA)is designed.Taking the road feel and the sensitivity of the steering as optimization objectives and the operation stability of the steering as a constraint,the system parameters are optimized.The simulation results show that the system optimized with MIGA can improve the steering road feel,and guarantee the operation stability and steering sensibility.

Optimizing Steering Angle Predictive Convolutional Neural Network for Autonomous Car

Deep learning techniques,particularly convolutional neural networks(CNNs),have exhibited remarkable performance in solving visionrelated problems,especially in unpredictable,dynamic,and challenging environments.In autonomous vehicles,imitation-learning-based steering angle prediction is viable due to the visual imagery comprehension of CNNs.In this regard,globally,researchers are currently focusing on the architectural design and optimization of the hyperparameters of CNNs to achieve the best results.Literature has proven the superiority of metaheuristic algorithms over the manual-tuning of CNNs.However,to the best of our knowledge,these techniques are yet to be applied to address the problem of imitationlearning-based steering angle prediction.Thus,in this study,we examine the application of the bat algorithm and particle swarm optimization algorithm for the optimization of the CNN model and its hyperparameters,which are employed to solve the steering angle prediction problem.To validate the performance of each hyperparameters’set and architectural parameters’set,we utilized the Udacity steering angle dataset and obtained the best results at the following hyperparameter set:optimizer,Adagrad;learning rate,0.0052;and nonlinear activation function,exponential linear unit.As per our findings,we determined that the deep learning models show better results but require more training epochs and time as compared to shallower ones.Results show the superiority of our approach in optimizing CNNs through metaheuristic algorithms as compared with the manual-tuning approach.Infield testing was also performed using the model trained with the optimal architecture,which we developed using our approach.

基于电动方向盘的拖拉机自动导航转向控制方法

为提高轮式拖拉机自动导航过程中转向控制的精度与稳定性,该研究以雷沃欧豹M704-2H拖拉机作为试验平台,采用电动方向盘作为转向执行机构,分析转向机械间隙对控制精度的影响,针对转向间隙特性设计转向控制算法。首先,为了获得准确的转向角,利用GNSS(global navigation satellite system)与二轮车模型快速标定虚拟轮转角,标定结果表明:虚拟轮转角的最大误差为1.3°,平均误差为0.11°。然后,对转向系统的机械间隙进行分析,设计一种带有间隙补偿的模糊PD(proportional derivative)转向控制算法,并在Simulink中验证算法的可行性。实车试验结果表明,该算法跟踪方波转角信号的响应时间为1.1 s,最大稳态误差为0.65°,平均稳态绝对误差为0.132°。跟踪正弦波转角信号的平均延时为0.5 s,最大误差为1.91°,平均绝对误差为1.09°。与无间隙补偿算法相比,有间隙补偿算法跟踪方波信号最大稳态误差减小了0.022°,平均稳态绝对误差减少了0.112°,角度误差在±0.2°内的时间提升了71%;跟踪正弦波信号最大误差减小了0.68°,平均绝对误差减小0.23°。田间直线导航转向控制试验结果表明,转角跟踪的绝对平均误差为0.61°,最大跟踪误差为2.82°,转向控制跟踪精度较高,稳定性好,满足导航作业需求。

NIM5驾驭氢钟守时能力分析

时间频率量值源头的独立自主是建设我国综合PNT体系的重要基础。为进一步评估基于我国国家秒长基准的原子时标守时能力,利用改进的有限脉冲响应钟差预测与频率调控方法,对激光冷却铯原子喷泉钟与氢钟的频差实测数据进行后处理,分别产生了自主型时标和溯源型时标。通过国际原子时合作链路将两个时标与协调世界时进行了长期比对实验,结果表明:2021年12月~2022年12月期间,两个时标的时间偏差均优于±4 ns、频率稳定度均优于8×10^(-16)/5 d。

基于二次转角约束的改进RRT路径规划算法研究

针对快速随机搜索树(RRT)算法存在节点扩展冗余、生成路径不满足车辆转角条件等问题,提出一种改进的二次转角约束RRT算法。首先,在传统RRT算法基础上对采样空间进行裁剪,引入目标导向策略减少采样时间;然后采用车辆膨胀处理和直线方法检测障碍物,并引入第一次转角约束得到粗解路径;接着对粗解路径建立二次转角约束并进行优化处理,获取优化路径后拟合,并进行仿真验证。结果表明,相比于引入目标导向策略的RRT算法,所提出的算法路径最大曲率降低了34.33%,平均曲率降低47.36%,扩展节点数降低47.62%,路径距离降低7.76%,规划时间缩短14.98%。

考虑履带机器人转向特性的全局路径规划

针对传统A星路径规划算法忽略车辆转向过程,导致规划路径行程时间较长的问题,提出了考虑履带车转向特性的改进A星算法。首先,扩大A星算法的搜索邻域,提高转向角的灵活度。其次,分析考虑履带车的转向特性,将转向时间加入到代价函数中,建立以时间最短为目标的代价函数,缩短履带车的行程时间。最后,构建删除冗余节点和新增优化节点的规则,改善路径的平滑性使规划的路径进一步优化。通过Matlab仿真和实车实验进行研究,研究结果表明:改进算法在路径长度、转向次数和行驶时间方面都有了减少,规划的路径要优于传统A星算法和对比文献算法。改进A星算法能有效提高所规划路径的质量,提升履带机器人的自主巡航能力和智能化水平。

基于粒子群优化算法PID参数优化的双电机耦合驱动履带车辆转向控制

针对电驱动履带车辆转向灵敏度不高、控制精度难的问题进行转向控制策略研究,对双电机耦合驱动履带车辆开展动力学分析,利用数学仿真软件建立转向动力学模型。设计一种基于粒子群优化算法对比例-微分-积分控制器(Proportional-Integral-Differential controller,PID)参数优化的转向控制策略,以改进型时间乘绝对误差积分指标为粒子群优化算法目标函数,对PID转向控制策略中的控制参数进行实时优化,动态调节控制参数,实现车辆转向控制系统优化输出。利用硬件在环仿真平台和实车试验,对控制策略进行仿真和实车试验验证,试验结果验证了控制策略的有效性。

基于凹凸性和转向角的古陶瓷碎片二次匹配算法

碎片拼接是古陶瓷修复的关键工作,针对古陶瓷碎片形状随机、数量大、表面纹理弱且存在局部缺损而导致算法的精度较低、匹配时间较长等问题,提出一种基于凹凸性和转向角的古陶瓷碎片二次配算法。在提取古陶瓷碎片轮廓曲线的基础上,通过先后使用粗匹配和细匹配的二次匹配组合实现碎片的两两精确匹配。一次粗匹配先通过多边形逼近碎片轮廓曲线,以降低轮廓的复杂性,再提取多边形的顶点凹凸性和顶点转向角构建一次轮廓特征集合,最后利用凹凸互补性和遍历顶点对齐的双模态特征初次匹配算法来寻找大致匹配段,并得到粗匹配点集。二次细匹配先随机选取粗匹配点集中的任意相邻两点点对来提取碎片轮廓片段,以减少轮廓点数量并提高算法效率,再计算轮廓片段的轮廓转向角以提取二次轮廓特征集合,最后利用基于粒子群优化的二次匹配来搜索精确匹配段,并得到细匹配点集。实验结果表明,该算法对二维古陶瓷碎片的拼接效果较好,且具有较强的鲁棒性,拼接误差不超过2%,运行时间效率相比已有算法提高了8%~20%。

基于细菌觅食-改进蚁群优化算法的水面无人船路径规划

为了解决水面无人船全局路径规划问题,提出了一种细菌觅食-改进蚁群优化算法(bacterial foraging-improved ant colony optimization algorithm,BF-IACOA)。相较于传统蚁群优化算法(ant colony optimization algorithm,ACOA),该算法在路径搜索策略上考虑水面无人船航行需要尽可能减少转向次数和完全规避过大转向角的约束,引入转向角启发因子,综合求解转移概率;同时引入细菌觅食算法的繁殖操作和趋化操作,改进信息素浓度的更新方式,解决传统ACOA容易陷入局部最优解和收敛速度较慢的问题。仿真结果表明,相较于传统ACOA,BF-IACOA的全局搜索能力得到较大幅度的提升,并且收敛迭代次数减少超过30%;在实际水域环境模型下,BF-IACOA可以通过14次迭代为无人船规划出全局可行路径。

机载一维相扫雷达空域稳定算法研究

传统上机载雷达是采用硬件稳定平台保持天线扫描空域稳定的,如今采用数据处理方法保持扫描空域电子稳定的研究正在引领发展趋势。文中针对机载平台运动引起天线波束指向角的改变以及由此带来的探测精度下降甚至目标丢失的问题,系统地探讨了一维相扫机载雷达保持扫描空域稳定的波束控制方法,建立了机载运动平台下指向角变换的数学模型,推导出由于平台运动引起的波束指向角变换公式,设计了一维相扫移相码的计算方法,通过仿真验证了计算精度、计算时间、波束指向精度和移相器位数选取等关键技术指标。结果表明空域稳定算法有效可靠,为机载相控阵雷达波束控制设计提供了通用化的设计思路,具有较大的工程意义和实用价值。

SMOTE数据预处理算法在砂型铸造复杂铸件缺陷预测中的应用

针对实际生产过程采集的复杂转向桥铸件工艺数据中冷隔、气孔、砂眼、缩孔等缺陷类别的数据量严重不平衡、复杂铸件缺陷预测模型准确率不高的问题,结合砂型铸造实际工况,引入了SMOTE(Synthetic Minority Oversampling Technique)数据预处理算法,探究其在砂型铸造复杂铸件缺陷预测中的应用。根据采集到的复杂铸件不平衡数据集的特点,基于SMOTE数据预处理算法,科学扩充了不平衡数据集,创建了可用于训练复杂铸件缺陷预测模型的平衡数据集,数据预处理前后的模型预测准确率从86.50%提高至97.91%。

基于线控底盘的无人驾驶轨迹跟踪控制研究

为提高结构化工况下无人驾驶汽车轨迹跟踪的精确性,提出一种融合纵向车速跟踪控制器的预瞄轨迹跟踪控制算法。针对四轮转向的特性,基于传统纯跟踪算法设计了四轮转向纯跟踪控制模型;针对纯跟踪算法跟踪效果受车速影响较大的缺陷,设计了纵向车速跟踪控制器并采用n维插值表方式在不同曲率处设置不同车速;将四轮转向纯跟踪算法融合纵向车速跟踪控制器,以达到更好的跟踪效果。仿真与实车试验表明,试验平台可较精确地跟踪规划轨迹并平稳抵达目标终点,实车轨迹跟踪过程中最大横向误差控制在0.497m。

基于FPGA的电动舵机滑模控制系统

电动舵机是现代无人机飞控系统的关键组成部分。针对传统电动舵机系统抑制干扰能力差、跟踪精度低、响应速度慢的问题,提出一种改进优化的自适应滑模控制(SMC)算法。首先,在进行系统建模和实验平台搭建时考虑了外界扰动和摩擦间隙等非线性问题,以此为基础设计了新的滑模面函数;其次,通过重新设计切换控制函数,使舵面偏转角度更加准确,有效的提高了系统的跟踪精度;最后,在系统中引入一阶低通滤波器,有效的消除了控制量抖振问题。并通过FPGA芯片实现了该算法,设计实验加以验证。实验结果表明,与传统PID算法相比,改进的滑模控制算法具有较强的抑制干扰能力和较快的动态响应速度,跟踪误差也有明显的下降。由此可见该算法在解决电动舵机的非线性问题方面具有较强的优势,基本满足舵面偏转的精准控制需求,在无人机产品的生产实践中具有较强的实际应用价值。

滑移转向无人车轨迹跟踪控制策略研究

针对滑移转向车辆构型的无人车,通过建立车辆动力学和运动学模型,对LQR(linear quadratic regulator)最优控制算法进行改进,在速度环采用一种自抗滑移的前馈控制,设计了一种轨迹跟踪控制策略。前馈控制有效提高了横摆角速度响应速度,可进一步优化LQR估计器的参数矩阵,保证车辆在提高轨迹跟踪控制精度的同时不影响跟随稳定性。采用Trucksim与Simulink联合仿真进行验证,结果表明:在双移线测试工况下,使用该控制策略的无人车最大横向误差比在双PID控制下小0.086 m,误差收敛速度2.1 s,稳态误差振荡幅值比滑模控制小约0.05 m;不同车速和不同附着系数路面下,无人车最大横向误差浮动小于0.02 m,误差收敛速度不变,设计算法控制效果稳定,验证了其鲁棒性。

基于改进天牛须算法的转向系统研究

农用自动转向系统要求较大的转向力矩、良好的可靠性等,而无刷直流电机具备低速大功率、启动转矩大、可靠性高、耐颠簸震动等特点,可以满足其要求。但传统PID控制的无刷直流电机存在响应慢、精度低等缺点,使其控制效果不甚人意。所以提出改进天牛须算法对其控制进行改进,通过建立电机模型,采用改进后的PID控制系统并完成仿真验证。经过仿真发现,经过改进天牛须算法优化后的转向控制系统具有良好的动、静态特性,转速相应快,抗干扰能力强,为农用转向系统研究提供有利的依据。