Combining Kalman filter and RLS-algorithm to improve a textile based sensor system in the presence of linear time-varying parameters

This paper presents an adaptive Kalman filter used as an observer in combination with a scaled least squares (LS) technique to improve a textile based sensor fusion. The focus of the application is to detect and monitor workplace particulate pollution. To control the sensor system around a reference current, a robust proportional-integral (PI) controller is used. In context of temperature variation, the sensor parameters resistance R and inductance L change in a linear way which is based on the linear range of the sensor characteristic. The adaption is performed with the help of an output-error (OE) model. The identification technique is based on the recursive least squares (RLS) method, which is used to estimate the parameters of the textile based model using input-output scaling factors. Through this proposed technique, a broader sampling rate and an input signal with low frequency can be used to identify the nano parameters characterizing the linear model. The simulation results emphasize that the proposed algorithm is effective and robust.