Optimized Artificial Neural Network Model and Compensator in Model Predictive Control for Anomaly Mitigation

Abstract

This paper presents a new artificial neural network (ANN)-based system model that concatenates an optimized artificial neural network (OANN) and a neural network compensator (NNC) in series to capture temporally varying system dynamics caused by slow-paced degradation/anomaly. The OANN comprises a complex, fully connected multilayer perceptron, trained offline using nominal, anomaly-free data, and remains unchanged during online operation. Its hyperparameters are selected using genetic algorithm-based meta-optimization. The compact NNC is updated continuously online using collected sensor data to capture the variations in system dynamics, rectify the OANN prediction, and eventually minimize the discrepancy between the OANN-predicted and actual response. The combined OANN-NNC model then reconfigures the model predictive control (MPC) online to alleviate disturbances. Through numerical simulation using an unmanned quadrotor as an example, the proposed model demonstrates salient capabilities to mitigate anomalies introduced to the system while maintaining control performance. We compare the OANN-NNC with other online modeling techniques (adaptive ANN and multi-network model), showing it outperforms them in reference tracking of altitude control by at least 0.5 m and yaw control by 1°. Moreover, its robustness is confirmed by the MPC consistency regardless of anomaly presence, eliminating the need for additional model management during online operation.

Model predictive trajectory tracking control of unmanned vehicles based on radial basis function neural network optimisation

“…Based on the development of deep learning theory, artificial neural networks (ANNs) 23,24 have become one of the most popular methods for describing complex nonlinear systems. 25 ANN-based MPC has been widely used in numerous applications, and trains its weights during operation to enhance prediction accuracy through model updates, 26 where a variety of ANN architectures have been examined. [27][28][29][30] These include recurrent neural networks (RNNs), multilayer perceptrons (MLPs) and radial basis function networks (RBFs).…”