Sensors 2001, 1, 38-52

In this paper a Kalman filter is used to cancel noise and interference contained in two accelerometers embedded in a car under performance tests. Here, the frequency bands of the signals of interest and their noise (deterministic as well as stochastic) are not strongly mixed with each other but it is very difficult to diminish the noise by using the classical approach to filtering. This reason, among others, justifies the necessity of the application of optimal filtering; and the Kalman filtering application allowed us to obtain optimal results (in the minimum mean-square sense) and signal-to-noise ratio improvements higher than 30 dB were achieved. The experimental results demonstrate the importance of using both analog signal conditioning and digital signal processing when dealing with signals corrupted by noise.