摘要
针对现有双轮自平衡结构机器人自动控制方法存在的倾角控制精度、机器人平衡性与工作效率相矛盾的不足,设计了一种基于四重优化模糊PID算法的机器人控制方法;先基于拉格朗日方程构建一种精准动力学模型,用于分析运动中机器人的平衡姿态,再基于线性优化降低模型的复杂度,在动力学模型优化过程中引入了附加扰动项实施动态补偿,进一步提升模型的性能;构建一种RBF神经网络模型并与自适应模糊算法、PID控制算法相融合,提升PID控制器的数据训练能力和最优结果输出能力,最后在神经网络输出层的3个输出节点控制上采用一种自适应调节机制,并通过最优控制函数获取到合理的控制结果;实验结果显示,提出四重优化融合方案下未出现超过25°倾角的情况,且在无障碍、有障碍和有坡度3种情况下,提出方法均未与障碍物碰撞,在有坡度和障碍物的情况下采样标准差的波动值仅为0.852。
Aiming at the shortcomings of the existing automatic control methods for two-wheel self-balancing structure robots,such as the contradiction between inclination control accuracy,robot balance and working efficiency,a robot control method based on the quadruple optimization fuzzy PID algorithm is designed.Firstly,a precise dynamic model is constructed based on the Lagrange equation to analyze the balanced posture of the robot in motion.Then,the complexity of the model is reduced through linear optimization.During the optimization process of the dynamic model,additional disturbance terms are introduced to implement dynamic compensation,further enhancing the performance of the model.An RBF neural network model was constructed and integrated with the adaptive fuzzy algorithm and PID control algorithm to enhance the data training ability and optimal result output ability of the PID controller.Finally,an adaptive adjustment mechanism was adopted in the control of the three output nodes of the neural network output layer,and reasonable control results were obtained through the optimal control function.The experimental results show that under the proposed quadruple optimization fusion scheme,there is no situation where the inclination Angle exceeds 25°.Moreover,in the three cases of no obstacles,with obstacles,and with slopes,the proposed method all involves collisions with obstacles.In the cases of slopes and obstacles,the fluctuation value of the standard deviation of the sampling site is only 0.852.
作者
刘婕
程卉
LIU Jie;CHENG Hui(Department of Information,Zibo Normal College,Zibo 255130,China;Library,Naval Aeronautical University,Yantai 264001,China)
出处
《计算机测量与控制》
2025年第12期161-166,236,共7页
Computer Measurement & Control
关键词
四重优化
双轮自平衡结构
RBF
PID控制
自适应调节机制
quadruple optimization
two-wheel self-balancing structure
RBF
PID control
adaptive regulation mechanism
