论文部分内容阅读
针对混凝土泵车支腿频繁发生开裂的问题,提出了一种基于支持向量机的故障诊断方法。混凝土泵车在施工过程中,施工状态的不同,直接影响到支腿的承载情况。对支腿故障进行分析,找出支腿最易出现开裂的关键位置,研究出了影响支腿关键位置承载的主特征量,搭建关键位置受力的支持向量机预测模型。运用MATLAB编写支持向量机预测程序,对模型进行训练和验证,通过输出的应力曲线图对支腿故障进行预测。实例验证了支持向量机对支腿关键位置应力预测的可行性。该方法相对于BP神经网络在小样本上更加精确,并为泵车工况参数的选择提供了理论支持。
Aiming at the problem that the outriggers of concrete pump truck frequently crack, a fault diagnosis method based on support vector machine is proposed. Concrete pump truck in the construction process, the different construction status, a direct impact on the bearing of the legs. The fault of the leg is analyzed and the most vulnerable position of the leg is found out. The main feature amount affecting the key position of the leg and the support vector machine prediction model of the key position are established. The support vector machine prediction program is written by MATLAB, the model is trained and verified, and the leg failure is predicted by the output stress curve. The feasibility of the support vector machine for stress prediction of the critical position of the outrigger is verified by an example. Compared with BP neural network, this method is more accurate in small samples and provides theoretical support for the selection of parameters of pump truck operation.