Power Prediction of Airborne Wind Energy Systems Using Multivariate Machine Learning

Abstract

Kites can be used to harvest wind energy at higher altitudes while using only a fraction of the material required for conventional wind turbines. In this work, we present the kite system of Kyushu University and demonstrate how experimental data can be used to train machine learning regression models. The system is designed for 7 kW traction power and comprises an inflatable wing with suspended kite control unit that is either tethered to a fixed ground anchor or to a towing vehicle to produce a controlled relative flow environment. A measurement unit was attached to the kite for data acquisition. To predict the generated tether force, we collected input–output samples from a set of well-designed experimental runs to act as our labeled training data in a supervised machine learning setting. We then identified a set of key input parameters which were found to be consistent with our sensitivity analysis using Pearson input–output correlation metrics. Finally, we designed and tested the accuracy of a neural network, among other multivariate regression models. The quality metrics of our models show great promise in accurately predicting the tether force for new input/feature combinations and potentially guide new designs for optimal power generation.

Towing Test Data Set of the Kyushu University Kite System

“…Using the data for a sensitivity analysis of the tether force with respect to design and operation parameters [19]. This work shows that the tow speed is the most important parameter affecting the tether force, which is consistent with the theoretical analysis.…”

“…The various test setups described in the literature all used steering mechanisms that were mounted on the towing vehicle. [19], the truck rendering is from http://freepik.com.…”

“…We have used the tow test data in a recent study on the use of machine learning algorithms for predicting the tether force [19]. The remainder of the paper is divided into three main sections.…”

“…Using the data for predicting the tether force employing multivariate machine learning algorithms [19]. In this work, the performances of different machine learning algorithms were assessed and the neural network model achieved the highest accuracy.…”

Reliability Analysis of a Home-scale Microgrid Based on a Threshold System

“…In their paper, a stochastic multiple-valued (SMV) approach is proposed to predict the reliability of two models of the system with non-repairable components and dynamically repairable components. The threshold-system model seems to be a very convenient model for exploring the reliability of renewable energy sources employing air-borne wind-energy vehicles [25][26][27][28].…”