The main obstacle against the penetration of wind power into the power grid is its high variability in terms of wind speed fluctuations. Accurate power forecasting, while making maintenance more efficient, leads to the profit maximisation of power traders, whether for a wind turbine or a wind farm. Machine learning (ML) models are recognised as an accurate and fast method of wind power prediction, but their accuracy depends on the selection of the correct hyperparameters. The incorrect choice of hyperparameters will make it impossible to extract the maximum performance of the ML models, which is attributed to the weakness of the forecasting models. This paper uses a novel optimisation algorithm to tune the long short-term memory (LSTM) model for short-term wind power forecasting. The proposed method improves the power prediction accuracy and accelerates the optimisation process. Historical power data of an offshore wind turbine in Scotland is utilised to validate the proposed method and compare its outcome with regular ML models tuned by grid search. The results revealed the significant effect of the optimisation algorithm on the forecasting models’ performance, with improvements of the RMSE of 7.89, 5.9, and 2.65 percent, compared to the persistence and conventional grid search-tuned Auto-Regressive Integrated Moving Average (ARIMA) and LSTM models.
In the industry field, the increasingly stringent requirements of lightweight structures are exposing the ultimately nonlinear nature of mechanical systems. This is extremely true for systems with moving parts and loose fixtures which show piecewise stiffness behaviours. Nevertheless, the numerical solution of systems with ideal piecewise mathematical characteristics is associated with time-consuming procedures and a high computational burden. Smoothing functions can conveniently simplify the mathematical form of such systems, but little research has been carried out to evaluate their effect on the mechanical response of multi-degree-of-freedom systems. To investigate this problem, a slightly damped mechanical two-degree-of-freedom system with soft piecewise constraints is studied via numerical continuation and numerical integration procedures. Sigmoid functions are adopted to approximate the constraints, and the effect of such approximation is explored by comparing the results of the approximate system with the ones of the ideal piecewise counter-part. The numerical results show that the sigmoid functions can correctly catch the very complex dynamics of the proposed system when both the above-mentioned techniques are adopted. Moreover, a reduction in the computational burden, as well as an increase in numerical robustness, is observed in the approximate case.
A complex fluid-structure interaction can often create nonlinear dynamic behaviour in the structure. This can be better estimated using nonlinear modal analysis, capable of identifying and quantifying the nonlinearity in the structure. In this study, the case of a vibrating beam submerged in liquid using a nonlinear parameter identification method is presented. This system is considered as an alternative propulsion mechanism, hence understanding the interaction between the fluid and the structure is necessary for its control. Here, impulse signals are used to characterise the numerical and experimental dynamics response of the system. Since the transient responses contain of a multi-component vibratory signals, a vibration decomposition method is used to separate the time response signals based on the dominant amplitude in the frequency response function. The separated time-series signals are then fitted to the nonlinear identification method to construct the backbone and damping curves. The modal parameters obtained from experimental data are then used as a base for the development of the analytical models. The analytical approaches are based on the Euler-Bernoulli beam theory with additional mass and quadratic damping functions to account for the presence of the fluid. Validations are carried out by comparing the dynamic responses of the analytical and experimental measurements demonstrating the accuracy of the model and hence, its suitability for control purposes.
25 October 2022 in International Journal of Mechanics and Materials in Design
Most of the optimisation studies of Vibration Energy Harvesters (VEHs) account for a single output quantity, e.g. frequency bandwidth or maximum power output, but this approach does not necessarily maximise the system efficiency. In those applications where VEHs are suitable sources of energy, to achieve optimal design it is important to consider all these performance indexes simultaneously. This paper proposes a robust and straightforward multi-objective optimisation framework for Vibration Piezoelectric Energy Harvesters (VPEHs), considering simultaneously the most crucial performance indexes, i.e., the maximum power output, efficiency, and frequency bandwidth. For the first time, a rigorous formulation of efficiency for Multi-Degree of Freedom (MDOF) VPEHs is here proposed, representing an extension of previous definitions. This formulation lends itself to the optimisation of FE and MDOF harvesters models. The optimisation procedure is demonstrated using a planar-shape harvester and validated against numerical results. The effects of changing some structural parameters on the harvester performance are investigated via sensitivity analysis. The results show that the proposed methodology can effectively optimise the global performance of the harvester, although this does not correspond to an improvement of every single index. Furthermore, the optimisation of each performance index individually results in a variety of design configurations that greatly differs from one another. It is here demonstrated that the design obtained with the multi-objective function here proposed is similar to the design obtained when optimising the efficiency.