Deformation Analysis And Prediction Based On Chaotic Time Series

Deformation analysis and forecast is a complicated system engineering,which involvesvarious theories and methods; It's a very important issue that how to introduce powerfulmathematical theory and methods of signal analysis to understand complexity and nonlinear of thedeformation, to extract the system information from the monitor deformation effectively, and toexplain and forecast the deformation. This paper based on researching the chaotic time series,discuss to analysis the deformation and predict from the angle of chaos theory, this paper mainachievements and specific content as follows:1. Introduces the basic concept of chaos and the development of chaotic time series,summarized the research status and progress of the chaotic time series in deformation analysis,and points out the existing problems, and puts forward the research contents of this paper.2. Introduces the identification method of chaotic time series, Based on the identification ofchaos ,analyse chaos feature of time sequence of deformation system,which can research systemall kinds of state,finding the characteristics of the movement, providing the necessary foundationto the next deformation analysis and prediction .3. Comprehensive introduce various technical means of chaos time series analysis andprediction, and put forward new methods of analysis and prediction, this paper introduces thetraditional analysis and forecasting which use dynamic system uncertainty and the nonlinearparameter ,studying with the method of using multiresolution analysis the system, and forecastingthe time series, studying the recent research of intelligent computing tools (such as radial basisfunction neural network, etc.), use their learning and iterative approximation ability to establishcomparatively complicated nonlinear analysis and prediction model. In view of the actualmeasurement noise, and introduces the wavelet denoising method of data.4. Study the correlation dimension and the Maximum Lyapunov index of deformationmonitoring data, demonstrates chaos characteristics exist in the deformation monitoring data.Prove theoretically discussed with the chaotic characteristics of deformation system, this paperdeduces the dynamic equation which has the Maximum Lyapunov index, and verify the deducefromthe example5. In the prediction of the chaos characteristics of the monitoring data, establish the forecastingmodel with the Maximum Lyapunov index, forecast with neural network and the multi-scaleforecast model combining wavelet analysis, from all kinds precision of forecasting , neuralnetwork adaptive ability is strong, it can not only fitting historical data well, also can accurately topredict system's future; While using the Maximum Lyapunov index that closely related with thechaotic systems developping forecast the system is very reasonable algorithm;wavelet multiscaleprediction algorithm proposed from multi-scale thoughts, merges with wavelet analysis,spectrumanalysis and neural network theory such algorithm theory, this algorithm can achieve preciseforecast fromthe result.