Composite control for biaxial traction-type full-scale fatigue testing of wind turbine blades

Full-scale blade fatigue testing requires accurate reproduction of the dominant fatigue response under actuator constraints. This study develops a composite controller for biaxial traction-type testing. A displacement-dependent map converts principal-axis force demand into non-negative cable tensions. Frequency-shaped LQR limits concentrated and high-frequency feedback corrections, while periodic amplitude compensation updates half-wave excitation from the preceding-cycle amplitude error. Simulink–Adams co-simulation gave a mean biaxial amplitude error of 0.54%, a mean NRMSE of 2.50%, and overshoot below 1% under nominal conditions. With 20% feedforward underestimation, PAC reduced flapwise and edgewise amplitude errors from 7.49% and 4.53% to 0.30% and 0.31%. Across R0–R4, mean amplitude errors did not exceed 0.72%. The controller gave smaller nominal amplitude errors than PID and the selected MPC benchmark. Its maximum measured execution time was 1.14 ms for a 25 ms control period.