基于BP神经网络的盾构机姿态与轨迹控制研究

为提高盾构机在隧道施工掘进中的运动轨迹精度,降低掘进中的"蛇行"量,提出一种基于BP神经网络的盾构机姿态与轨迹控制参数补偿方法.首先,对影响盾构机姿态与轨迹的参数进行分析,从而结合动量定理和动量矩定理,构建盾构机空间6自由度运动的动力学方程.其次,利用BP神经网络反向学习功能,对盾构机姿态控制参数进行预测,并借助专家知识系统的推理功能,对推进系统姿态进行调整,从而使盾构机更好地沿着预定的设计轴线掘进.最后,对新算法进行仿真,验证了算法的可行性.

Research on the attitude and trajectory control ofshield machine based on BP neural network

In order to improve the shield machine for the tunnel excavation trajectory precision and reduce excavation "hunting",it introduces the control parameter compensation method for shield machine attitude and trajectory based on BP neural network.It analyzes the affecting parameters of shield machine attitude and trajectory,establishes the motion equations of shield machine dynamics in the space of six degrees of freedom combining with the theorem of momentum and moment of momentum theorem,applies BP neural network to reverse learning function,predicts the posture parameters on step motion control and makes better shield function along a predetermined track driving.At last,the algorithm is simulated and the feasibility of the algorithm is verified.