Papers by Keyword: ARX Model

Paper TitlePage

Authors: Wen Qiao, Guo Ming Liu, Jin Wen He
Abstract: Research on the dynamic characteristics of gravity dam was carried out by adopting ARX model using the seismic response data of concrete gravity dam at Shui-Kou hydropower station. The applicability and effectiveness of single-output-multiple-input ARX model were deduced and verified. A corresponding computer program was developed, and performed to identify the modal parameters of the system. The identified natural frequencies and damping ratios were basically same with the results by traditional peak point pick-up method, and also close to the finite element method (FEM) results. It is indicated that the structure natural frequencies and damping ratios are determined by the characteristics of the structure, the dynamic characteristics identified by ARX model are correct, and ARX model can avoid frequency leakiness when smoothing processing and Fourier transform are conducted in solving process of the peak point pick-up method. The modal identification can be applied to other structures.
Authors: Ming Chih Huang, Yen Po Wang, Chien Liang Lee
Abstract: In this study, damage localization of frame structures from seismic acceleration responses is explored using the DLV technique and ARX model for system identification. The concept of the DLV method is to identify the members with zero stress under some specific loading patterns derived by interrogating the changes in flexibility matrix of the structure before and after the damage state. Success of the DLV method for damage localization lies on the ability to identify the flexibility matrix. The ARX model, a discrete-time non-parametric auto-regressive system identification technique is adopted to identify the modal parameters (natural frequencies, transfer functions and mode shapes) from which the flexibility matrices of the intact and damaged structures are constructed. To explore the effectiveness of the DLV method, a five-storey steel model frame with diagonal bracings was considered for seismic shaking table tests. The damage conditions of the structure were simulated by partially removing some of the diagonals. With the flexibility matrices of both the intact and damaged structures synthesized on a truncated modal basis, the damage locations have been successfully identified by the DLV method for either single or multiple damage conditions, regardless of the damage locations. This study confirms the potential of the DLV method in the detection of local damages from global seismic response data for frame structures.
Authors: Jing Shu Wang, Ming Chi Feng, Chang An Zhu
Abstract: In order to realize the characteristics of the positioning stage, the system identification has been developed. The location data has been interpolated and normalized so as to improve the accuracy of system identification. After data preprocessing, the identification based on ARX model and the least square method has been carried out. The results indicate that the ARX model is appropriate to describe the characteristics of positioning system.
Authors: Huai Yuan Liu, Jian Hua He, Song Chen
Abstract: Based on the principle of the system identification, combined Simulink with System Identification Toolbox from MATLAB, the least square estimation method is selected to establish a system of ARX model, and Akaike Information Criterion (AIC) was used in the identification of model order, compared with the original model to study the fitting accuracy, and the validity of the model is examined by residual analysis. This approach overcomes the disadvantages of the complexity and difficulty in traditional programming model. Compared to other program identification method, it has a short modeling time, and it is clear, reliable, intuitive visual, good scalability. Furthermore, the model parameters, result and system can be easily modified, assessed and verified. This method of system modeling and simulation can be used for reference to aerospace and other fields.
Authors: Bo Chen, Ji Cai Feng
Abstract: The exploration of marine resources is becoming more and more important, underwater weld technology is an important technology for the exploration. However, the underwater weld technology is very immature; the underwater weld quality is not as good as the offshore welding. To control and improve the underwater weld quality automatically, arc and visual sensors were used to obtain the information reflecting the weld quality, and ARX (AutoRegressive eXogenous) model was use to model the process, experiment result showed that the prediction results of the model were accurate enough for automatically controlling the process.
Authors: Narongkorn Dernlugkam, Parkpoom Chokchairungroj
Abstract: This paper presents the system identification of a prototype of hydraulic platform compare with two methods between the Autoregressive model structure with exogenous input and the Lagrange’s equation with SolidWorks’ simulation analysis tool. The system identification was identified, then the model was employed to simulate the multi-sine input response and the simulation results were compared to the experimental results. It was found that the model was developed using Lagrange’s equation more accurate than the Autoregressive model structure that having percentage best fit of 93.45% for the azimuth angle control and 92.25% for the elevation angle control. Overall, this method could be accepted with a good percentage best fit that compared with experiment results.
Authors: Dong Jing Yang, Jin Wu Gao, Le Lun Jiang, Tan Xiao
Abstract: Thrombelastograph device (TEG) is a measuring instrument of blood viscoelastic properties during coagulation. The measuring temperature of TEG is fixed at 37oC while in some surgery cases, lower temperature surroundings may be adopted. Therefore a new type of TEG with a controllable themostatic system has been designed to mimic various temperature surroundings in surgery. In this paper, a small-sized high accuracy thermostatic system for TEG was designed and its system identification was built to facilitate the development of control strategy. ARX model was supposed to analyze the system identification of the thermostatic system by Matlab System Identification Toolbox. Residual analysis method was adopted to verify the identified model. The results showed that the simulation data of ARX model was consistence with the measured data (matching degree was about 93%). Transfer function of the system can be applied to develop its control strategy.
Authors: Hong Xia Pan, Li Ming Sun
Abstract: The EMD method as a means of signal pre-processing is used to decompose the vibration signal of gearbox into a number of IMF components, using the largest energy component as an effective component of the output response signal and the input shaft rotate speed or torque signal as the system input signal, then the time-series ARX model is established, in this paper. Extracted model parameters are as a feature vector of recognized working state of gearbox, and as input parameters of RBF neural network to achieve automatic recognition of faults in the gearbox. The results show that: ARX model can enhance the anti-jamming capability of model, and have great significance for improving the accuracy of fault diagnosis by calculating the fluctuations of input signal into the model.
Showing 1 to 8 of 8 Paper Titles