Improving the Real-Time Response of an ADXL202 Accelerometer Placed in a Car Under Performance Tests by using Adaptive Filtering

In this paper, a recursive least-squares lattice (RLSL) adaptive filter was used to estimate the relevant signal coming from an ADXL202 accelerometer placed in car under performance tests. In this practical application, the signal of interest was buried in a broad-band noise background where we had little knowledge of the noise characteristics. The results of the experiment were satisfactory and, in order to compare the type of conventional filters used in today´s cars to cancel the noise corrupting the information coming from automotive sensors against optimal adaptive filters, the signal coming from the ADXL202 was also filtered by using several lowpass digital Butterworth filters. The results of the experiment showed that the signal-to-noise (SNR) ratio improvement achieved by using the RLSL adaptive filter was 42.66 times better than the best SNR improvement that was achieved by using the classical filters.