基于PSO-LSSVM的塔机起升机构虚拟传感器研究  

Research on virtual sensor of tower machine liftmechanism based on PSO-LSSVM

作　　者：孙兴旺 殷晨波[1] 乔文华 SUN Xingwang;YIN Chenbo;QIAO Wenhua(School of Mechanical and Power Engineering,Nanjing University of Technology,Jiangsu Nanjing 211816,China)

机构地区：[1]南京工业大学机械与动力工程学院,江苏南京211816

出　　处：《工业仪表与自动化装置》2023年第2期60-64,共5页Industrial Instrumentation & Automation

摘　　要：目前塔机起升机构的安全监测多采用布置传感器的方法,载荷测量方式主要是通过在定滑轮处加装轴销式称重传感器,高度测量多采用安装外接编码器的方法。此方法安装、更换成本高且布线发杂,因此该文提出了一种通过变频器输出数据来计算起升载荷以及吊钩高度虚拟传感测量方法。首先介绍了起升机构的组成,通过变频器中的电机编码器数据完成了吊钩高度的换算。接着描述了载荷测量虚拟传感的原理与方法,选用基于粒子群优化的最小二乘支持向量机算法来建立吊重测量模型,通过对数据的关联性分析确定了模型的输入输出。最后通过现场试验验证了载荷测量虚拟传感方法的可行性,且吊重测量误差不超过4%。At present,the safety monitoring of the tower crane lifting mechanism mostly adopts the method of arranging sensors.The load measurement method is mainly through installing the axle pin weighing sensor at the fixed pulley,and the height measurement mostly adopts the method of installing the external encoder.This method has high installation and replacement cost and miscellaneous wiring distribution,so this paper proposes to calculate the lifting load and the hook height virtual sensing measurement method.Firstly,the height of the lifting mechanism is introduced by the motor encoder data in the frequency converter.The principle and method of load measurement virtual sensing are described.PSO-LSSVM is selected to establish the lifting weight measurement model,and the input and output of the model are determined through the correlation analysis of the data.Finally,the feasibility of the virtual sensing method of load measurement is verified by the field test,and the lifting weight measurement error does not exceed 4%.

关 键 词：塔式起重机 起升载荷测量 转矩电流 虚拟传感 粒子群优化最小二乘支持向量机 

分 类 号：TH213.3[机械工程—机械制造及自动化]