Time series forecasting for multidimensional telemetry data based on Generative Adversarial Network in a Digital Twin

Abstract

The research related to Digital Twins has been increasing in recent years. Besides the mirroring of the physical word into the digital, there is the need of providing services related to the data collected and transferred to the virtual world. One of these services is the forecasting of physical part future behavior, that could lead to applications, like preventing harmful events or designing improvements to get better performance. One strategy used to predict any system operation is the use of time series models like Autoregressive Integrated Moving Average (ARIMA) or Long-Short Term Memory (LSTM) and improvements implemented using these algorithms. Recently, deep learning techniques based on generative models such as Generative Adversarial Networks (GANs) have been proposed to create time series, and the use of LSTM has gained more relevance in time series forecasting, but both have limitations that restrict the forecasting results. Another issue found in the literature is the challenge of handling multivariate environments/applications in time series generation. Therefore, new methods need to be studied in order to fill these gaps and, consequently, provide better resources for creating useful Digital Twins. In the proposed method we introduce the integration of a Bidirectional LSTM (BiLSTM) layer with a time series obtained by GAN that leads to improved forecasting of all feature of the available dataset in terms of accuracy. The obtained results demonstrate improved prediction performance.