System Identification Of Motion Control System On Placement Machine Based On RBF Network

Placement machine is key equipment of SMT product line, which is also common in electronics manufacturing industry. The performance of placement machine mainly depends on the motion control system's speed, accuracy and stability. While developed countries has nearly monopoly the core technology, researching on it in China has big significance.On the basis of using an old useless Quad QSP-2 equipment machine, exchanging the primary equipment such as industrial computer and motion controller, constructing a new motion control system which has the basic functions of control, this paper centers on the system identification of motion control system on placement machine, by which characteristics of the object can be known well and good foundation for realizing better control strategy can be supplied. Outlines of the renovation plan includes:(1)Identify modal structure of motion control system, using the least-squares algorithm;(2)Identify the mathematic modal of motion control system, using RBF network;(3)Complete system identification of motion control system based on RBF network, using the structure got from the identification based on the least-squares algorithm;For more details, works have been done as follows:(1)Designed the data collecting program;(2)Expounded the system identification based on least-squares algorithm and RBF network, both of which were used in actual identification;(3)Compared the identification accuracy based on different methods;(4)Optimized the control software and modified motion mode which can solve the problem existed;(5)Recorded implementation process, laying solid foundation for further researches.