A novel numerical integrator for structural health monitoring

Accelerometers are one of the most commonly used sensor in structural control and health monitoring applications. However the accelerometer output is polluted with noise and bias signals. Obtaining velocity and position from these acceleration measurements are not trivial. Direct integration will result in an output drift. In this paper baseline correction and filtering techniques are used to overcome these problems. Experimental results on a linear actuator illustrates the effectiveness of the proposed method.