摘要
针对麦克纳姆轮农业机器人在智能大棚中的路径跟踪问题,建立了运动学模型和动力学模型,设计了一种新型的双环比例微分-分数阶比例积分导数(Proportional Derivative-Fractional Order Proportional-Integral Derivative,PD-FOPID)控制器对全局路径进行动态跟踪控制。对于控制器参数多且整定困难的问题,首先采用帐篷映射初始化种群策略、精英主义记忆策略、动态余弦权重策略和柯西—高斯变异策略对原始野马算法进行改进,然后利用略改进野马算法(Improved Wild Horse Optimizer,IWHO)对控制器最优增益参数优化。实验结果表明:所开发的算法在探索和开发阶段方面性能优异,且PD-FOPID控制器在整定工作中表现突出。路径跟踪仿真证明,设计的双环PD-FOPID控制器比FOPID控制器更具显著的优势,能够避免动态误差累积,快速响应调整到规划路径,在提高农业大棚机器人路径跟踪控制质量方面具有巨大的潜力。
Aiming at the path tracking problem of Mecanum-wheel Agriculture Robot(MAR)in the intelligent shed,we first establish the Kinematic and kinetic models,and design a new two-loop Proportional Derivative-Fractional Order Proportional-Integral Derivative(PD-FOPID)controller for the global path dynamic tracking control,for the problem of many controller parameters and the difficulty of rectification,we first use the tent mapping initialization population strategy,elitist memory strategy,dynamic cosine weighting strategy,and Corsi-Gaussian variation strategy to improve the original wild horse algorithm,and then use the slightly Improved Wild Horse Optimizer(IWHO)to optimize the optimal gain parameters of the controller,the experimental results show that the developed algorithm has excellent performance in the exploration,development stages and the PD-FOPID.The experimental results show that the developed algorithm performs well in the exploration and development stages and excels in the tuning of the controller.According to the path-tracking simulation,it is proved that the two-loop PD-FOPID controller designed in this paper has significant advantages over the FOPID controller in terms of avoiding dynamic error accumulation and fast response adjustment to the planned path.
作者
穆占海
艾尔肯·亥木都拉
郑威强
Mu Zhanhai;Aierken·Haimudula;Zheng Weiqiang(College of Smart Manufacturing Modern Industrial,Xinjiang University,Urumqi 830047,China)
出处
《农机化研究》
北大核心
2025年第2期1-8,18,共9页
Journal of Agricultural Mechanization Research
基金
国家自然科学基金项目(52265039)。
关键词
麦克纳姆轮农业机器人
路径跟踪
运动学模型
动力学模型
新型双环控制器
改进野马算法
mecanum-wheel agriculture robots
path tracking
kinematic models
kinetic models
new type of double-loop controller
improved wild horse optimizer
