A new approach for error minimization of piezoelectric sensor output variations using fuzzy logic

Performance of electronic sensors used for physical measurements is highly depends upon the operating conditions. Environmental changes e.g., temperature variations do affect the sensor outputs especially for those which are used for very precise measurement. The effect of temperature variation from ambient to the harsh drastically degrades the electronic sensor output performance e.g., piezoelectric sensor in our case. It is very hard to model the output variation of piezoelectric sensor whose output is non-deterministic or random in nature and it also change with temperature. Normal mathematical modelling techniques to correct error are very difficult to implement due to complexity and nonlinearity. Data has been acquired from piezoelectric sensor used for pressure measurement whose output variation is random in nature with respect to the temperature variations. The scope of this paper is to study the behaviour of piezoelectric sensor output signal taking into consideration the different varying temperature profiles. Artificial intelligence technique like Fuzzy is used to minimize the error up to desirable amount. The novelty in our work is that Fuzzy approach has been implemented on this kind of engineering problem. Results have been compared with mathematical technique i.e., curve fitting based upon the error correction criterion. Fuzzy approach has shown significant improvement in error minimization of piezoelectric sensor as compared to the mathematical techniques.