A Novel Prognostic Model of the Degradation Malfunction Combining a Dynamic Updated-ARIMA and Multivariate Isolation Forest: Application to Radar Transmitter

In the prognosis of radar transmitter degradation malfunction, there are some restrictions, such as fact that it is difficult to obtain fault samples and monitoring data cannot reach threshold. For these a novel data-driven prognostic method proposed predict in which dynamic updated-auto-regressive integrated moving average be used subsequent time-step microwave measurement historical data, multivariate isolation forest established without detect malfunction. The validity portability model verified using two-type malfunction experiments. experimental results show can predicted at least 10 time-steps (100 min) before occurrence Compared with existing prediction methods, needs less no samples, artificial thresholds, extracting features. This complete when relevant restrictions.