Abstract:The availability of accurate empirical models for multi-step-ahead (MS) prediction is desirable in many areas. Some ANN technologies, such as multiple-neural-network, time-delay neural network (TDNN), and adaptive time-delay neural network (ATNN), have proven successful in addressing various complicated problems. The purpose of this study was to investigate the applicability of neural network MS predictive model. Motivated by the above mentioned technologies, we proposed a hybrid neural network model which integrated characteristics decomposition units, and a dynamic spline interpolation unit into the multiple ATNNs. Inside the net, the regular and certain information were extracted to ATNN, while both time delays and weights were dynamically adapted. The yearly average of the sunspots, which has been considered by geophysicists, environment scientists, and climatologists as a complicated nonlinear system, was selected to test hybrid model. Comparative results were presented between traditional MS predictive model based on TDNN and the proposed model. Validation studies indicated that the proposed model is quite effective in MS prediction, especially for single factor time series.

No abstract

This is the author's copy of the publication as archived with the DLR's electronic library at http://elib.dlr.de. Please consult the original publication for citation.

This paper proposes a method for lifelong learning of Recurrent Neural Networks, such as NNARX, ESN, LSTM, and GRU, to be used as plant models in control system synthesis. The problem is significant because in many practical applications it is required to adapt the model when new information is available and/or the system undergoes changes, without the need to store an increasing amount of data as time proceeds. Indeed, in this context, many problems arise, such as the well known Catastrophic Forgetting and Capacity Saturation ones. We propose an adaptation algorithm inspired by Moving Horizon Estimators, deriving conditions for its convergence. The described method is applied to a simulated chemical plant, already adopted as a challenging benchmark in the existing literature. The main results achieved are discussed.

No abstract