脐带缆拉弯组合疲劳试验机恒拉力控制

Constant tension control of umbilical cable tension-bending combined fatigue testing machine

由于深海脐带缆的变拉伸刚度特性、卧式加载条件下脐带缆过长、疲劳试验机弯曲端伸缩状态切换而引入冲击噪声等因素,导致恒拉力控制难度高、精度低。在对疲劳试验机控制系统研究的基础上,将模型参考自适应控制算法应用于该试验机的恒拉力控制系统中。针对液压系统中存在的非线性时变参数,提出了自适应线性神经网络与归一化最小均值M估计(ADALINE-NLMM)的自适应控制策略。其利用系统估计的输出误差调整自适应的神经网络的权值,同时利用最小均值M估计算法调整系统中的不确定参数。根据液压系统内部频率变化而跟踪参考模型的输出,削弱脉冲噪声的干扰,提高了控制系统的鲁棒性。不同弯曲角度下脐带缆的静态拉伸试验表明:系统的静态跟踪误差最大不超过3%,平均跟踪误差接近0.3%。一定角度范围内动态拉伸试验表明,脐带缆拉伸端施加恒定的拉力的控制误差不超过10%。结果表明:提出的模型具有良好的恒拉力控制精度和鲁棒性。

The variable tensile stiffness characteristics of the deep-sea umbilical cable,the excessive length of the umbilical cable under horizontal loading conditions,and the introduction of impact noise due to the switching of the bending end expansion state of the fatigue testing machine lead to high difficulty and low accuracy of constant tensile force control.Based on the study of the control system of the fatigue testing machine,the model reference adaptive control algorithm is applied to system.Aiming at the nonlinear time-varying parameters in the hydraulic system,an adaptive control strategy and the normalized minimum mean M estimation(ADALINE-NLMM)was proposed.It adjusts the weight of the adaptive neural network by using the output error estimated by the system.It adjusts the uncertain parameters by using the least-mean M estimation algorithm.The output of the reference model is tracked according to the change of the internal frequency of the hydraulic system,which weakens the interference of impulse noise and improves the robustness.Static tensile tests show that the maximum static tracking error of the system is less than 3%,and the average tracking error is close to 0.3%.The dynamic tensile test shows that the error of constant tensile force applied by the system to the tensile end of the umbilical cable is less than 10%.The results show good accuracy of constant tension control and strong robustness.