Comparison of Compensation Algorithms for Smart Sensors With Approach to Real-Time or Dynamic Applications

This paper presents the dynamic assessment and the comparison of four compensation algorithms programmed in commercial hardware with a restricted computational capability. For each of the four algorithms implemented in a low cost microcontroller embedded in a data acquisition system, the frequency response was evaluated to find out their performance in real-time or dynamic environments. The data acquisition system is an analysis tool proposed to model the compensation process as the first-order system. With the transfer functions in s and z domains and some experimental analysis, the dynamic features for all the four compensation schemes were estimated. The development of modern measurement systems implies the use of hardware and software; they require evaluation beyond the number of samples per second. Then, it is necessary to determine their capabilities and performance for real-time or dynamic applications, because their development is in part based on the physical features and the computational capability of the hardware used. Also, smart sensors are used in the development of measurement systems, and an important capability is the compensation to eliminate the systematic errors and ease the sensor calibration. Available literature lacks information regarding to the performance of compensation algorithms in real-time or dynamic applications. Then, the analysis process exposed in this paper can be a useful methodology to the designer or the end user to evaluate the performance for dynamic or real-time conditions of any measurement system. The frequency and timing parameters would change according to the algorithm implemented and processor used in the data acquisition system.