- 12
- 0
- 约5.86千字
- 约 12页
- 2018-10-09 发布于福建
- 举报
桥吊实时工况统计分析和预测
桥吊实时工况统计分析和预测
摘 要:为准确掌握桥式起重机(简称桥吊)的动态信息,对上海外高桥集装箱桥吊的实时工况进行统计分析与预测.通过分析桥吊起升电机的振动信号和温度信号统计特性,得到振动与温度的关系;通过对信号数据的预处理,建立支持向量机(Support Vector Machine, SVM)训练模型,其中对惩罚参数和核函数参数采用交叉验证的方法进行优化;利用得到的训练模型预测后继的振动信号.与单纯用振动信号或温度信号所建的模型相比,这种振动信号与温度信号相结合的模型对电机振动信号预测的准确性更高.
关键词:桥式起重机; 实时工况; 统计分析; 预测; 支持向量机
中图分类号:TH215; TP18; TP274文献标志码:A
Statistical analysis and prediction on
real-time operating condition of bridge crane
XU Xiuzhong, JIANG Shan
(Logistics Engineering College, Shanghai Maritime Univ., Shanghai 201306, China)
Abstract:In order to grasp the dynamic information of bridge crane accurately, the statistical analysis and prediction on real-time working condition of container bridge crane in Waigaoqiao port in Shanghai is conducted. Through the statistical feature analysis on vibration signal and temperature signal of lifting motor of bridge crane, the relationship between vibration and temperature is obtained; through preprocessing the data of vibration signal and temperature signal, the Support Vector Machine (SVM) model is established, and the penalty parameter and kernel function parameter are optimized by cross validation. The model is then used to predict the vibration signal. Compared with the models which are established according to vibration signal or temperature signal respectively, the model which combines vibration signal and temperature signal provides more accurate prediction of the vibration signal of the lifting motor.
Key words:bridge crane; real-time working condition; statistical analysis; prediction; support vector machine
0 引 言
桥式起重机(简称桥吊)的工作状态关系到港口起重机械安全生产和生产效益.桥吊的工况特殊,特别是电机部分经常连续工作,负荷较大,因而容易出现问题.如果能对电机的工作状态进行有效预测,准确掌握起重机械的动态信息,就可以避免电机出现较大问题、延误生产,从而减少不必要的经济损失.
目前,预测桥吊电机工作状态的方法较多,但是采用这些方法预测通常需要大量样本,而且泛化能力差.支持向量机(Support Vector Machine,SVM)是专门针对有限样本情况的,目标是得到现有信息下的最优解而不仅仅是样本数趋于无穷大时的最优解,即得到的是全局最优点,这一方法可以解决神经网络方法中无法避免的局部极值问题.
1 电机实时工况统计特征分析
1.1 电机实时工作数据的特点
当大梁起升时,电机不工作,没有振动信号;当大梁放平工作时,由于载荷对电机产生的冲击使电机产生
原创力文档

文档评论(0)