摘要
Accurate fault prediction can obviously reduce cost and decrease the probability of accidents so as to improve the performance of the system testing and maintenance. Traditional fault prediction methods are always offline that are not suitable for online and real-time processing. For the complicated nonlinear and non-stationary time series, it is hard to achieve exact predicting result with single models such as support vector regression （SVR）, artifieial neural network （ANN）, and autoregressive moving average （ARMA）. Combined with the accurate online support vector regression （AOSVR） algorithm and ARMA model, a new online approach is presented to forecast fault with time series prediction. The fault trend feature can be extracted by the AOSVR with global kernel for general fault modes. Moreover, its prediction residual that represents the local high-frequency components is synchronously revised and compensated by the sliding time window ARMA model. Fault prediction with combined AOSVR and ARMA can be realized better than with the single one. Experiments on Tennessee Eastman process fault data show the new method is practical and effective.
Accurate fault prediction can obviously reduce cost and decrease the probability of accidents so as to improve the performance of the system testing and maintenance. Traditional fault prediction methods are always offline that are not suitable for online and real-time processing. For the complicated nonlinear and non-stationary time series, it is hard to achieve exact predicting result with single models such as support vector regression （SVR）, artifieial neural network （ANN）, and autoregressive moving average （ARMA）. Combined with the accurate online support vector regression （AOSVR） algorithm and ARMA model, a new online approach is presented to forecast fault with time series prediction. The fault trend feature can be extracted by the AOSVR with global kernel for general fault modes. Moreover, its prediction residual that represents the local high-frequency components is synchronously revised and compensated by the sliding time window ARMA model. Fault prediction with combined AOSVR and ARMA can be realized better than with the single one. Experiments on Tennessee Eastman process fault data show the new method is practical and effective.
