Kite System Research Articles

Dynamic Characterization, Flow Modeling, and Hierarchical Control of an Energy-Harvesting Underwater Kite in Realistic Ocean Conditions

This paper presents a hierarchical control framework for a kite-based marine hydrokinetic (MHK) system, along with a detailed characterization of the dynamic and energetic performance of the system under realistic flow conditions. The underwater kite, which is designed to be deployed off of an offshore floating platform, features a closed-loop controller that executes power-augmenting, cyclic cross-current flight. The robustness of the kite's undersea flight control algorithm is demonstrated in a realistic four-dimensional flow model (which captures both low-and high-frequency spatiotemporal variations in the current) that accounts for turbulence and wave effects, which is coupled with a detailed dynamic model that captures the six-degree-of-freedom kite and floating platform dynamics, in addition to the tether dynamics. Using data obtained by the Coastal Data Information Program (CDIP) 192 Oregon Inlet buoy [1], wave data from the Wave Information Studies Hindcast model [2], and a spectral turbulence model developed at Florida Atlantic University, we demonstrate the robustness of the kite's control system and the sensitivity of both average net power output and peak-to-average power to wave parameters. In common wave conditions, the average and net power output are shown to be highly robust to the peak period and significant wave height. In extreme wave conditions, the peak-to-average power ratio is shown to be highly positively correlated with an effective wave energy density metric, which characterizes the wave energy density presented to the kite system based on a weighted distribution along depth of the kite.

Integrated physical design, control design, and site selection for an underwater energy-harvesting kite system

This paper presents a co-design framework that optimizes the kite design, site, and controller of a kite-based marine hydrokinetic (MHK) energy-harvesting system. The formulation seeks to maximize a techno-economic metric, namely power-to-mass ratio, by simultaneously considering three key categories of decision variables while accounting for the coupling between the three. The simultaneous consideration presents computational challenges associated with optimizing a large number of decision variables, a subset of which (control variables) are time trajectories. The multi-fidelity co-design formulation presented in this work utilizes two techniques, namely nesting and layering, to solve the optimization problem in a computationally tractable manner without significantly compromising on accuracy. Specifically, nesting allows for efficient integration of the three optimization sub-modules into one integrated framework without accuracy losses, whereas layering allows for successive design space reduction as the overall optimization progresses from using a low-fidelity model to using a higher-fidelity model. The resulting integrated co-design tool was applied to a region of interest off the North Carolina coast to optimally choose a combination of deployment site, kite design, and control strategy. We show that the integrated co-design tool results in a two-fold performance improvement over benchmarks derived from sequential (or independent) optimization of the kite categories, thereby underscoring the need for co-design. Computational effectiveness is demonstrated by comparing the computational cost of the nested and layered approach against the estimated computational costs that would be required to perform a single high-fidelity integrated optimization over the entire design space.

Theoretical analysis of the power generation of pumping cycle kite power systems compared to traditional wind turbines in Aberdeen

The United Kingdom (UK) has pledged to reduce its greenhouse-gas emissions to net zero by 2050. However, in July 2022, the High Court ruled that the government’s net zero strategy failed to outline policies that would enable it to meet the target. As a result, the government published a new plan that sets out opportunities for innovative technologies like Floating Offshore Wind Manufacturing, and hydrogen, which will not only help us reach net zero. For wind, the government’s goal is to develop up to 50 GW of offshore wind by 2030. The challenge of such an expansion is huge. It means the United Kingdom will need to install an estimated 3,200 new, and much larger, wind turbines by 2030—roughly three new turbines every 2 days. Airborne Wind Energy (AWE) systems are part of a new class of wind energy converters that is receiving considerable attention in the renewable energy generation arena. AWE systems benefit from the steadier and stronger wind streams at high altitudes to generate more energy from wind while avoiding the expense of tower construction. This paper presents a feasibility study of using AWE systems in the UK and particularly in the City of Aberdeen to generate renewable energy. The characteristics of wind energy distributions were theoretically investigated by developing a wind speed distribution model, and then the annual power production of a kite system and a turbine system with 30 kW generator were analysed by applying the annual wind profile in Aberdeen to the performance models of these two systems. It was found that the annual power production of the kite system was two times higher than that of a normal turbine system.

Stability and Performance of an Undersea Kite Operating in a Turbulent Flow Field

In this article, we examine the effects of flow disturbances resulting from turbulence on the dynamic behavior of an underwater energy-harvesting kite system that executes periodic figure-8 flight. Due to the periodic nature of the kite’s operation, we begin by assessing orbital stability using the Floquet analysis and stroboscopic intersection analysis of a Poincaré section, with the former analysis performed on a simplified “unifoil” model and the latter performed on a six-degree-of-freedom (6-DOF)/flexible tether model. With periodic stability established, a frequency-domain analysis based on a linearization about the kite’s path is used to predict the quality of flight path tracking as a function of the turbulence frequency. To validate the accuracy of these simulation-based predictions under flow disturbances, we compare the predictions of the kite’s behavior against the results of small-scale tow testing experiments performed in a controlled pool environment.

Recursive Gaussian Process-Based Adaptive Control of a Ducted Kite System for Tidal Energy Harvesting

In this brief, we introduce an adaptive control approach for a novel tidal energy harvesting device called the Duct-Sail. The design combines features of kite-based energy-harvesting devices and ducted turbines to realize power augmentation through cross-current flight at low flow speeds along with the flow augmentation, stationary performance potential, and protection offered by a duct. At low flow speeds, the Duct-Sail executes high-speed figure-8 cross-current flight to generate rated power. At higher flow speeds, it curtails its cross-current motion to limit structural loading while still delivering rated power. We present a detailed dynamic model of the system along with design parameters for an initial prototype. We also present model-based and nonmodel-based adaptive control strategies that are used to control the intensity of cross-current flight in a time-varying flow profile. Lastly, we present simulation results using the real flow data from a candidate installation site, which enables a practically meaningful comparison of various adaptive control strategies.

Prediction regions based on dissimilarity functions

This paper presents a new methodology to obtain prediction regions of the output of a dynamical system. The proposed approach uses stored past outputs of the system and it is entirely data-based. Only two hyperparameters are necessary to apply the proposed methodology. These scalars are chosen so that the size of the obtained regions is minimized while fulfilling the desired empirical probability in a validation set. In this paper, methods to optimally estimate both hyperparameters are provided. The provided prediction regions are convex and checking if a given point belongs to a computed prediction region amounts to solving a convex optimization problem. Also, approximation methods to build ellipsoidal prediction regions are provided. These approximations are useful when explicit descriptions of the regions are necessary. Finally, some numerical examples and comparisons for the case of a non-linear uncertain kite system are provided to prove the effectiveness of the proposed methodology.

Design and Model Identification of a Power Kite Wind Energy System

Power kite energy systems use power kites attached to the ground by tethers that transfer the wind mechanical power to ground electrical generators. In order to remain in the air, the power kite should fly continuously in “lying” eight form trajectories. One key to the optimal operation of this system is automatic control of the kite flight, requiring a reliable real-time position measurement and advanced closed-loop control of the position and speed of the kite during its flight. The basis of a good controller is the identification and modelling of the dynamic system. This is the purpose of the current work. A power kite system prototype is built and identified on the basis of input-output data. Validation of the obtained model is also provided.

Analysis of the work performance of the lift mode crosswind kite power system based on aerodynamic parameters

High-altitude wind power generation mainly collects wind energy at a certain altitude through tethered aircraft. This article discusses the current status of the research on the power system of the cut wind flying kite, and derives the relationship between the drag coefficient and the solidity factor and aerodynamic coefficient χ of the towed kite in the lift mode. Furthermore, the formulas of lift-type working power coefficient and waste power coefficient are discussed. It is pointed out that the power coefficient CPLk will increase with the increase of the aerodynamic coefficient χ and increase with the decrease of the Solidity factor σ, but the drag coefficient will be a≤1/2. The product of the Solidity factor and the aerodynamic coefficient σχ≤4. Analyzed the characteristics of each system parameter, and then used the concepts of effective power and waste power to derive the efficiency formula of the system, and pointed out that the range of reasonable work efficiency is 0.33∼0.5.

Integrated Plant, Site, and Control System Co-Design for an Underwater Energy-Harvesting Kite System

Integrated Plant, Site, and Control System Co-Design for an Underwater Energy-Harvesting Kite System

Laboratory Simulation of Electricity Harvesting from an Underwater Kite

This paper addresses the problem of experimentally characterizing the energy conversion losses associated with electricity harvesting from tethered marine hydrokinetic kites. The authors present a laboratory setup intended for this characterization. The setup utilizes two geared brushless DC electric machines. The first electric machine is intended for energy harvesting from a scaled kite system in a lake environment. The second electric machine is used in the laboratory for emulating the mechanical loads associated with energy harvesting, thereby allowing the setup to operate as a hardware-in-the-loop testing system. Experiments conducted using this system, for a kite flight trajectory optimized for mechanical power harvesting, reveal a significant discrepancy between the mechanical power seen at the tether release drum versus the electrical power ultimately generated. Simple linear regression shows that much of this discrepancy can be attributed to mechanical friction within the setup (including the electric machine's gearing), as opposed to electrical impedance. This creates a strong motivation for additional research on trajectory optimization for electrical power generation.

Generalized aerodynamic models for crosswind kite power systems

This paper presents two novel semi-analytical models for predicting the aerodynamic performance of crosswind kite power systems (CKPSs), where the kite induction effects on the oncoming flow are taken into account. The blade element momentum theory forms the backbone of the models. The effects of reel-out ratio, solidity factor, rotor incidence angle, side-slip angle and tether drag are included in the formulation for the axial induction factor and power output. For simplicity, the wake rotation and the tangential induction factor are neglected. Aerodynamic model 1 is developed for predicting the reel-out power with uniform inflow assumption, and it is suitable for CKPSs with ground-based power generation. Aerodynamic model 2, on the other hand, can predict both the reel-out and torque powers with either uniform or non-uniform inflow assumption, and the model can be used for CKPSs with ground-based, or on-board power generation, or with the combination of the two. Some parametric studies have been conducted for a generic kite system with pre-defined aerodynamic efficiency parameters to highlight the effects of incidence and side-slip angles. In addition, a particular CKPS and its variants are examined to show the individual and combined effects of incidence angle, side-slip angle, tether drag and airfoil shape on the induction factor and power output.

Dynamic Response Analysis of Composite Tether Structure to Airborne Wind Energy System under Impulsive Load

The tether structure plays the role of transferring the traction force of an airborne wind energy system (AWES) to the fixed or mobile ground system with less motion and maintains the flying airborne system as a critical component. The implementation of a geometrically tailored tether design in an AWES could avoid unwanted snap-through failure, which can be take place in a conventional tether structure under impulsive loading. This concept relies on the redundant load path of the composite structure composed of tailored length and strength. In this study, the dynamic response of this composite tether structure to airborne wind energy systems, such as a kite wind power system, was analytically investigated. Also, for very long tether applications, an approximate model of the tether response was developed, which resulted in a dramatic reduction of computational efforts while preserving the accuracy quite well compared to the exact solution.

System Identification of a 6 m2 Kite Power System in Fixed-Tether Length Operation

The traction force of a kite system can be utilized for extracting energy from high-altitude wind. This paper discusses a system identification algorithm derived to obtain real-time governing equations for the kite system based on real-flight data, obtained from a 6 m2 kite power system developed in Kyushu University as an airborne wind energy project. The paper presents the system set-up, the design, the experimental results, a system identification algorithm, and the parameters identified for the kite used. The current stage of the project considers the kite application as a fixed-tether length system with a ground kite control unit. The control strategy is designed to work as a hardware-in the-loop to keep receiving the data and controlling the kite in real time. The experimental tests employed are divided into four distinct ones, and the data of the kite’s attitude, position, and tension forces are recorded. The tension forces resulted from these tests are presented for different wind speeds and flight modes. Ultimately, a novel system identification algorithm that evaluates the correlation between the tension force and the kite’s rolling angle over the four tests is applied, thereby enabling to study the kite behavior as a preliminary step for the achievement of autonomous flight.

Lab-Scale, Closed-Loop Experimental Characterization, Model Refinement, and Validation of a Hydrokinetic Energy-Harvesting Ocean Kite

Abstract This paper presents a study wherein we experimentally characterize the dynamics and control system of a lab-scale ocean kite, and then refine, validate, and extrapolate this model for use in a full-scale system. Ocean kite systems, which harvest tidal and ocean current resources through high-efficiency cross-current motion, enable energy extraction with an order of magnitude less material (and cost) than stationary systems with the same rated power output. However, an ocean kite represents a nascent technology that is characterized by relatively complex dynamics and requires sophisticated control algorithms. In order to characterize the dynamics and control of ocean kite systems rapidly, at a relatively low cost, the authors have developed a lab-scale, closed-loop prototyping environment for characterizing tethered systems, whereby 3D printed systems are tethered and flown in a water channel environment. While this system has been shown to be capable of yielding similar dynamic characteristics to some full-scale systems, there are also fundamental limitations to the geometric scales and flow speeds within the water channel environment, making many other real-world scenarios impossible to replicate from the standpoint of dynamic similarity. To address these scenarios, we show how the lab-scale framework is used to refine and validate a scalable dynamic model of a tethered system, which can then be extrapolated to full-scale operation. In this work, we present an extensive case study of this model refinement, validation, and extrapolation on an ocean kite system intended for operation in the Gulf Stream or similar current environments.

Towing Test Data Set of the Kyushu University Kite System

Kites can be used to harvest wind energy with substantially lower material and environmental footprints and a higher capacity factor than conventional wind turbines. In this paper, we present measurement data from seven individual tow tests with the kite system developed by Kyushu University. This system was designed for 7 kW traction power and comprises an inflatable wing of 6 m2 surface area with a suspended kite control unit that is towed on a relatively short tether of 0.4 m by a truck driving at constant speed along a straight runway. To produce a controlled relative flow environment, the experiment was conducted only when the background wind speed was negligible. We recorded the time-series of 11 different sensor values acquired on the kite, the control unit and the truck. The measured data can be used to assess the effects of the towing speed, the flight mode and the lengths of the control lines on the tether force.

Power Prediction of Airborne Wind Energy Systems Using Multivariate Machine Learning

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.

Adaptive Flight Path Control of Airborne Wind Energy Systems

In this paper, we applied a system identification algorithm and an adaptive controller to a simple kite system model to simulate crosswind flight maneuvers for airborne wind energy harvesting. The purpose of the system identification algorithm was to handle uncertainties related to a fluctuating wind speed and shape deformations of the tethered membrane wing. Using a pole placement controller, we determined the required locations of the closed-loop poles and enforced them by adapting the control gains in real time. We compared the path-following performance of the proposed approach with a classical proportional-integral-derivative (PID) controller using the same system model. The capability of the system identification algorithm to recognize sudden changes in the dynamic model or the wind conditions, and the ability of the controller to stabilize the system in the presence of such changes were confirmed. Furthermore, the system identification algorithm was used to determine the parameters of a kite with variable-length tether on the basis of data that were recorded during a physical flight test of a 20 kW kite power system. The system identification algorithm was executed in real time, and significant changes were observed in the parameters of the dynamic model, which strongly affect the resulting response.

Optimal Cyclic Spooling Control for Kite-Based Energy Systems

This paper presents a control strategy for optimizing the the spooling speeds of tethered energy harvesting systems that generate energy through cyclic spooling motions which consist of high-tension spool-out and low-tension spool-in. Specifically, we fuse continuous-time optimal control tools, including Pontryagin’s Maximum Principle, with an iteration domain co-state correction, to develop an optimal spooling controller for energy extraction. In this work, we focus our simulation results specifically on an ocean kite system where the goal is to optimize the spooling profile while remaining at a consistent operating depth and corresponding average tether length. This paper demonstrates a 14-45% improvement (depending on the operating tether length and environmental flow speed) in power generation compared to a baseline, heuristic, control strategy.

Hierarchical Control Design and Performance Assessment of an Ocean Kite in a Turbulent Flow Environment

This paper presents a hierarchical control framework for a kite-based marine hydrokinetic (MHK) system that executes power-augmenting cross-current flight, along with simulation results based on a high-fidelity turbulent flow model that is representative of flow conditions in the Gulf Stream. The hierarchical controller is used to robustly regulate both the kite’s flight path and the intra-cycle spooling behavior, which is ultimately used to realize net positive energy production at a base station motor/generator system. Two configurations are examined in this paper: one in which the kite is suspended from a surface-mounted platform, and another in which the kite is deployed from the seabed. To evaluate the robustness of this control framework in a realistic ocean environment, we present simulation results whereby we superimpose low-frequency data from the Mid Atlantic Bight South Atlantic Bight Regional Ocean Modeling System and acoustic Doppler current profiler measurements with a high-frequency turbulence model, resulting in a high-fidelity 3D spatiotemporal flow field that is presented to the kite system. Based on this simulation framework, we demonstrate the effectiveness of the control system both in terms of robust flight and power generation.

A new perspective on the aerodynamic performance and power limit of crosswind kite systems

In this paper, a new perspective on the aerodynamic performance modelling of crosswind kite power systems (CKPSs) is provided, where the effects of the induction factor or flow retardation by the kite are taken into account. For simplicity, only CKPSs in straight downwind configuration are considered, where the kites sweep an area perpendicular to the wind direction. Moreover, the in-plane or tangential induction factor is neglected. It is argued that the concept of the swept-area-normalized power coefficient, which is commonly used for conventional wind turbines, is not practically important for CKPSs. Instead, a kite-area-normalized power coefficient concept is adopted and is shown to be a more appropriate metric of performance of kite systems. The kite-area-normalized useful and wasted power coefficients for both a lift and a drag mode CKPS are plotted at different values of solidity factor and aerodynamic efficiency. Moreover, it is shown that the two modes of power generation, i.e. lift and drag modes, yield the same amount of useful power when the kite system solidity factor is essentially zero – this is in agreement with the underlying assumption in the crosswind kite power principle. For non-zero solidity factors, thus non-zero induction factors, however, our results suggest that the drag mode has higher power generation potential than the lift mode.