基于正交试验和神经网络的液压杆稳定性研究

以Euler公式为基础,发现动力猫道液压杆的稳定性与其缺陷的形状、面积、深度及位置有关。通过对液压杆屈曲载荷的模拟计算,运用正交试验法分析缺陷的形状、面积、深度及位置对液压杆屈曲载荷的影响。研究结果表明:缺陷的面积、位置和深度对液压杆稳定性具有显著影响,其发生不显著情况的概率分别为0.007, 0.021, 0.008;缺陷的形状对液压杆稳定性影响显著性相对较小,其发生不显著情况的概率为0.123;对液压杆稳定性影响大小次序依次为缺陷面积>缺陷深度>缺陷位置>缺陷形状。正交试验建立的神经网络预测模型经实验验证,对含缺陷液压杆屈曲载荷预测具有很高的准确度,此方法可为含缺陷液压杆的稳定性校核及安全使用提供有力的技术保障。

Stability of Hydraulic Rod Based on Orthogonal Test and Neural Network

ased on the Euler formula, it is found that the buckling load of the hydraulic rod used in the power catwalk is related to the shape, area, depth and position of the defect. Through the simulation calculation of hydraulic rod buckling load, the influence of defect shape, area, depth and position on hydraulic rod buckling load was analyzed by orthogonal test method, and the neural network prediction model of hydraulic rod buckling load was established by orthogonal test data. The results show that the area, position and depth of defects have significant influence on the stability of the hydraulic rod, and the probability of the non-significant situation in one test is 0.007, 0.021 and 0.008. The shape of the defect has a relatively small influence on the stability of the hydraulic rod, and the probability of the non-significant situation in one test is 0.123. Therefore, the order of influence on the stability of hydraulic rod is as follows: defect area >defect depth >defect position >defect shape. Moreover, the neural network prediction model established by orthogonal test data has been verified to be very accurate in predicting the buckling load of the hydraulic rod, which can provide powerful technical support for the checking and safe use of the hydraulic rod with defects.