Machine learning of electro‑hydraulic motor dynamics

摘要

In this paper we propose an innovative machine learning approach to the hydraulic motor load balancing probleminvolving intelligent optimisation and neural networks. Two different nonlinear artificial neural network approaches areinvestigated, and their accuracy is compared to that of a linearised analytical model. The first neural network approachuses a multi-layer perceptron to reproduce the load simulator dynamics. The multi-layer perceptron is trained using theRprop algorithm. The second approach uses a hybrid scheme featuring an analytical model to represent the main systembehaviour, and a multi-layer perceptron to reproduce unmodelled nonlinear terms. Four techniques are tested for theoptimisation of the parameters of the analytical model: random search, an evolutionary algorithm, particle swarm optimisation,and the Bees Algorithm. Experimental tests on 4500 real data samples from an electro-hydraulic load simulatorrig reveal that the accuracy of the hybrid and the neural network models is comparable, and significantly superior to theaccuracy of the analytical model. The results of the optimisation procedures suggest also that the inferior performanceof the analytical model is likely due to the non-negligible magnitude of the unmodelled nonlinearities, rather thansuboptimal setting of the parameters. Despite its limitations, the analytical linear model performs comparably to thestate-of-the-art in the literature, whilst the neural and hybrid approaches compare favourably.