Theories, Methods And Application For Improving The Accuracy Of Intelligent Test System

Intelligent test is an important way to intelligent control, which is animportant branch of intelligent control. Accuracy is one of the most important ability targets for the system of intelligent test, consequently, research in improving the accuracy of the system of intelligent test becomes one of the hot points of inland and external research.Nonlinear error, Repeated error and Temperature drift error are three major factors in the accuracy of test system, so in this paper, the research status on overcoming the three errors and their shortage are introduced. Furthermore, an improved subsection correction method basing on BP neural network for reducing nonlinear error, a data fusion method basing on RBF neural network for reducing temperature drift error and a digital filter method for reducing repeated error are proposed as the emphasis of this paper, in which, the principle, improved model of neural network and simulation experiments of the three methods are described in detail.Finally, a synthesis verifier system for spiracle with high accuracy is put into practice by applying the three methods, which includes a synthesis study module basing on neural network for reducing nonlinear error and temperature drift error, an intelligent measure module, a zero adjustment module and a data processing module. Experiments and application indicate that the methods proposed in this paper is efficient and practical for reducing errors of test system.