Study On Model-based Rapid Testing Method Of Equipments

Following with the rapid increase on complexity of the new weapons, the current testing technology is hard to satisfy the requirements of the rapid development of equipments. On the one hand, the requirement of standardization, virtualization and modeling to the testing technology becomes more and more strong. On the other hand, more and more new techniques are introduced into the testing field and form testing capability. This thesis mainly investigated a rapid testing capability building method--the model-based rapid testing method of equipments, which can be used to build the mathematical models and physical models of equipments according to the observed data when their priori information is limited, thus the effective test can be implemented according to the physical models of the equipments.The development history and key techniques of automatic test systems (ATS) home and aboard are introduced at the beginning of this thesis. The typical products of ATS, especially the next ATS frame--NxTestATS are analyzed, which is proposed by American army in 2002. Using the signal model and the signal classification defined in IEEE 1641 Standard, the building method of the mathematical model element library and physical model element library is proposed to ensure the unity of signals both in theory and in form. According to the observed data of equipments, a system identification method based on gene expression programming is proposed and the corresponding software program is designed. This algorithm can be used to identify the structure and parameters of a system at the same time and has the ability to restrain the white noise and colored noise in observed data. After the mathematical model of the equipment is obtained, the differential evolution algorithm is adopted to implement model matching according to the model library built in advance. That is to say, to get the physical model of the equipment rapidly using the mapping between the mathematical model elements and physical model elements, thus the fast fault diagnosis and position can be implemented. Furthermore, the concept and identification method of the system signal model (SSM) is proposed, which integrates the advantages of mathematical model and physical model. It can be used to build the test requirements of the unit under test (UUT) rapidly, establish the models of the UUT and ATS themselves, and exploit the transplantable testing program. At the end of this thesis, several experiments are designed to verify the effectiveness of the model-based rapid testing method of equipments. It provides the technical supports for the virtual test based on software simulation and can be used to form testing capability for new equipments rapidly.