A Degradation Fault Prognostic Method of Radar Transmitter Combining Multivariate Long Short-Term Memory Network and Multivariate Gaussian Distribution

Abstract

In the prognosis of radar transmitter degradation fault, there are some problems, such as the total sample size and fault sample size of sensor monitoring data are small, and the monitoring data can not reach the fault threshold. To solve these problems, a prediction model combining the multivariate long short-term memory networks with multivariate Gaussian distribution is proposed, in which the long short-term memory networks predict the subsequent time step of multi-sensor monitoring data, and the multivariate Gaussian distribution model constructed by few fault samples is used to realize the prediction of degraded faults. The key parameters of the model are determined by the data processing experiments of double monitoring points, and the validity of the model is verified by the data processing experiments of multiple monitoring points. The experimental results show that the degradation fault can be predicted effectively within 10% of the total time series of monitoring data. Compared with the traditional radar warning after failure, the model can effectively predict the degradation fault of the transmitter when the fault sample size is low and the fault threshold is not reached.INDEX TERMS Multivariate long short-term memory network, multivariate gaussian distribution, prognostic method, degradation fault of radar transmitter.

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

“…Partial magnifications of the prediction result and RMSE are shown in Figure 7. It can be seen from the figure that the RMSE of the two microwave measurement points predicted by DU-ARIMA are 0.068333 and 0.0077417, respectively, which are much lower than the 0.15848 and 0.028292 in [21]. The experiment results show that the DU-ARIMA has higher a prediction accuracy.…”

“…Table 6 compares the proposed degradation malfunction prognostic model with the previous method. It can be seen from the table that, compared with [19][20][21], the method proposed in this paper has a smaller sample size. It also does not need to set the artificial threshold, extract the feature, and use the fault samples for training.…”

“…The research to date has tended to focus on prognostic methods based on measurement and data-driven models, which have been widely used in transportation [1][2][3], biology [4,5], machinery [6][7][8], energy [9][10][11], market [12][13][14], radar [15][16][17][18][19][20][21][22][23][24], and other applications. Among them, the research results of malfunction prognoses in radar-related fields continue to emerge.…”

“…Li et al [20] used a diagnosis model with a dynamic threshold based on deep learning to predict the meteorological radar fault. Zhai et al [21] proposed a prediction model combining the multivariate long short-term memory (LSTM) networks with a multivariate Gaussian distribution to predict the transmitter degradation fault. Wang et al [22] developed a datadriven fault-detection method using principal component analysis to detect the fault of the fans in a radar cooling system.…”

“…Reference [19,20,23,24] also needed extensive historical data and fault samples for its training sets. Although reference [21] solves the two restrictions of radar transmitter degradation fault, this method is not accurate enough for data prediction, and at least one fault sample is still needed.…”