Because wind power is connected to the grid on a large scale, frequency fluctuation in the power grid, which is defined as a system safety risk to the power grid, occurs from time to time. According to the grid code rules of China, wind turbines are required to be equipped with primary frequency modulation or inertia response control capability, which are used to support the safe and stable operation of the power grid. During the traditional frequency modulation process of the wind turbine, power limiting operation or pitch angle reservation is generally adopted to ensure that the reserved energy can be released at any time to support the frequency change in the power grid. However, the frequency support method leads to a large loss of power generation, and does not consider the coordination between mechanical load characteristics control and primary frequency modulation. In this paper, a mechanical load optimization control strategy for a wind turbine during the primary frequency modulation process, based on LIDAR (light detection and ranging) feed forward control technology, is proposed and verified. Through LIDAR feed forward control, the characteristics of incoming wind speed can be sensed in advance, with the consequence that the wind turbine can participate in, and actively control, the primary frequency modulation procedure. According to the characteristics of incoming wind, for instance the amplitude and turbulence, simultaneously, the size of the reserved pitch angle can be adjusted in real time. This kind of method, coordinating with the mechanical load of the wind turbine, can be used to reduce both the ultimate load and fatigue damage as much as possible. Finally, the mechanical load characteristics of the wind turbine with and without the control strategy are compared and studied through simulation. The research results show that the load optimization control strategy based on LIDAR feed-forward control technology can effectively reduce the fatigue and ultimate loads of the wind turbine while supporting the frequency change in the power grid; especially for the fatigue load of tower base tilt and roll bending moments, the reducing proportion will be about 4.3% and 6.3%, respectively.

Compressive residual stresses generated during thin film deposition may lead to undesirable film damage, such as delamination, buckling, and flaking, ultimately leading to the failure of the device employing the film. Understanding the residual stress generation and role in these damage mechanisms is necessary to preserve thin film integrity and optimize its functional properties. Thin shell theory has been used for decades to predict buckling but the results have not yet been correlated with experimental data since the techniques used to measure stress in metallic films were not able to do so at the required micron scale until recently. Micro scanning X-ray diffraction now enables the direct mapping of the local stress of metallic films. In this paper, finite element method based on thin shell theory and synchrotron X-ray micro diffraction have been used to determine stress maps of thin film buckling patterns. Calculations of the stress distribution in the metallic films have been performed taking into account the buckling geometry determined from optical measurements. Stress distributions over gold blisters and tungsten wrinkles obtained with the two techniques are in fair agreement and allow for the accurate determination of the stress relaxation profile from the bottom to the top of the buckling, validating the thin shell theory model.

Amid the global energy transition, the rapid growth of wind turbine deployment has highlighted the need for accurate fatigue load prediction to support structural design and ensure operational reliability. This study proposes a neural network-based method for estimating fatigue loads at critical locations of large wind turbines. Wind speed, turbulence intensity, and yaw angle were used as input features, while the damage equivalent loads at the blade root, tower base, and yaw bearing served as prediction targets. A dataset comprising 2139 operating conditions was constructed, and two predictive models—an artificial neural network (ANN) and a Bayesian neural network (BNN)—were developed and evaluated using standard error metrics. The results show that the BNN consistently achieves lower prediction errors and higher goodness-of-fit values than the ANN across all outputs, demonstrating improved accuracy and stability. The BNN model attained excellent predictive performance, with an overall coefficient of determination (R²) of 0.9998, a root mean square error (RMSE) of 0.012, and a mean absolute percentage error (MAPE) of only 0.1877%. These findings indicate that probabilistic neural networks hold strong potential for enhancing fatigue load prediction and can provide valuable support for wind turbine structural assessment, design optimization, and active yaw control strategies.