Optimal Placement of Blade Tip Timing Sensors Considering Multi-mode Vibration Using Evolutionary Algorithms

Blade tip timing (BTT) as a noncontact measurement technique is developed to monitor the health condition of aero-engine rotor blades. The circumferential placement of BTT sensors plays an important part in identifying blade vibration parameters. If the placement is suboptimal, the reconstructed displacement signal can be sensitive to the measurement noise while certain useful information is lost. This paper presents an optimal placement of blade tip timing sensors considering multi-mode vibration using evolutionary algorithms. The installation angle of BTT sensors around the casing is taken as the design variable and the inaccessible angle of the casing is taken as the constraint condition. Particle swarm optimization (PSO) algorithm is the preferred optimization one in this work due to its strong robustness and fast convergence. The condition number of the design matrix of blade multi-mode vibration reconstruction model is taken as the fitness function of PSO, the minimum of which corresponds to the optimal sensor positions. Simulation experiment results show that compared with the random placement of BTT sensors, the more blade modes required to be identified along with the circumferential Fourier fit algorithm employed, the more prominent the advantage of the proposed method is.