No abstract

As the sizes of wind turbine blades increase, it is becoming more difficult for designing fatigue tests to meet the requirements of blades' certification. In order to design accurate fatigue tests for large-scale wind turbine blades, this paper proposes an improved optimization based on particle swarm optimization. The advantages of this optimization algorithm are its accuracy in predicting test moments and its ability in matching distributions of test moments. The shell model, which can model some special details of large-scale blades, is used to predict test moments instead of the beam model in the optimization algorithm. Besides, removing the blade tip is introduced in to enrich test methods, which enlarges the dimensions of variables in the optimization algorithm. Furthermore, moment errors are measured in both the absolute type and the relative type. Based on the study of these two types of moment errors, the objective function and constraints are designed to guide the decrease and distribution of moment errors. At last, edge-tests of a 38-m blade (shell and beam model), a 60-m blade (beam model), and a 38.75-m blade (beam model) are designed by this improved optimization, and the optimal test moments and distributions of moment errors prove the abilities of this improved optimization. Differences of the moment errors and test durations between the results designed by the standard testing method and those designed by the improved testing method demonstrate the advantages of improved optimization.