The purpose of this paper is to analyze the modern deep neural networks such as nonlinear autoregressive network with external inputs and a recurrent neural network called long short-term memory for wind speed forecast for long-term and use the prediction for fatigue analysis of a large 5 MW wind turbine blade made of composite materials. The use of machine learning algorithms of advanced neural network applied for engineering problems is increasing recently. The present paper therefore brings as important connection between these latest machine learning methods and engineering analysis of complex wind turbine blades which are also the focus of researchers in renewable system design and analysis. First, a nonlinear autoregressive network with external inputs neural network model using Levenberg–Marquardt back propagation feed forward algorithm is developed with 5 years of environment parameters as input. Similarly, a long short-term memory based model is developed and compared. The chosen long short-term memory model is used for developing two-year wind speed forecast. This wind pattern is used to create time varying loads on blade sections and cross-verified with National Renewable Energy Laboratory tools. A high-fidelity CAD model of the NREL 5 MW blade is developed and the fatigue analysis of the blade is carried out using the stress life approach with load ratio based on cohesive zone modeling. The blade is found to have available life of about 23.6 years. Thus, an integrated methodology is developed involving high-fidelity modeling of the composite material blade, wind speed forecasting using multiple environmental parameters using latest deep learning methods for machine learning, dynamic wind load calculation, and fatigue analysis for National Renewable Energy Laboratory blade.

The use of advance composite materials is increasing in various industrial applications such as renewable energy, transportation, medical devices, etc. As the demand for stability under high mechanical, thermal, electrical and combined loads is increasing, research is being focused on developing newer types of composites and developing analytical and numerical methods to study composite plates as well. The present work is aimed to provide a comprehensive review of research in the structural analysis of composite plates along-with research trends in the last 15 years. The article first presents the evolution of plate theories comparing their formulations, applicability and discusses some key papers, results and conclusions. Evolution of research from the equivalent shear deformation theories (ESL) such as first order theory and higher order theories based on various shape strain functions e.g., polynomial, trigonometric to layer-wise, zigzag and displacement potential theories is presented. The comparative analysis of various solution approaches is done based on a review of research work in the structural analysis of plates. This is followed by review of meshless analysis methods for composite materials highlighting problem domains where conventional finite element analysis (FEA) approach has limitations. This article also presents a discussion on the new methods of plate analysis such as region-by-region modeling, hierarchic modeling and mixed FE and neural network based modeling. An attempt has been done in this article to focus on research trends in the last 15 years.

The purpose of this article is to analyze various wind speed-forecasting methods, select the appropriate method for developing synthetic wind speed for 1-year period at Salem in Tamilnadu state in India, and use it for the structural and fatigue analysis of a small horizontal-axis wind turbine blade made of composite material. Various forecasting models such as Markov chain, Kalman filter, and autoregressive integrated moving average are evaluated, and a long-term wind speed pattern at Salem is developed using Markov chain. This wind pattern is used to create time-varying loads using the blade element momentum on blade sections. Then, the fatigue analysis of the blade is carried out using the stress life approach. The blade is found to have available life of about 20 years and the critical area for fatigue is found on the skin near the root of the blade. Various blade skin materials are also compared for fatigue performance. A cohesive zone model of the adhesively joined root joint is also developed and analyzed for fatigue at the metal-composite joint. Thus, an integrated methodology involving high-fidelity modeling of the blade, wind forecasting, and static and fatigue analysis is developed for horizontal-axis wind turbine blade for locations where historically wind speed measurements are available for short time.