dsPIC based piecewise linear interpolation for sensor applications

摘要

Industrial applications of sensing and measurement with non-linear sensor output distorted by noise require linearization processes. Among various linearization techniques Piecewise linearization is found suitable because of its faster execution speed using a simple look-up table and it significantly reduces the program and RAM memory requirement. The paper presents Microcontroller with on chip digital signal processor dsPIC30F6014 based piecewise linearization for sensor-based applications. To implement the technique, the thermocouple sensor is used to measure temperature and interfaced with Microchip's dsPIC30F6014 processor kit under MPLAB IDE development tool. Availability of software tools such as real-time operating systems, TCP/IP, and prepackaged DSP algorithms has further enhanced the usability of such architectures in a wide extent of sensor applications. Further, the advantage of DSP and microcontroller core together can be used for finite impulse response (FIR) or infinite impulse response (IIR) filtering to remove signal noise or computation of a fast Fourier transform (FFT) in processing and normalization of input measured data. Finally, result obtained with thermocouple for temperature measurement is presented with accuracy of +/- 1% of full scale (0-200℃ ). The developed system is cost effective and supports both RS232 and Wireless interface for connecting to host/wireless multi sensor network applications.