Ali Haghighat Mesbahi - Department of Chemical Engineering, Polymer Engineering Group, Isfahan University of Technology, Isfahan۸۴۱۵۶-۸۳۱۱, Iran
Elham Khakrah
Saied Nouri Khorasani - Department of Chemical Engineering, Polymer Engineering Group, Isfahan University of Technology, Isfahan۸۴۱۵۶-۸۳۱۱, Iran
Jalal Barzin - Department of Mathematic and Industry, Isfahan University of Technology, Isfahan, Iran

Expert system or in more common expression, fuzzy if-then rule systems try to present relation between significance and precision via mathematical functions. In fact, using knowledge of an expert in formulating an exact output is the main purpose of this knowledge based systems. The object of present study is investigating Expert system parameters on modeling performance of input/output system of release behavior from an injectable in situ forming Drug Delivery System. These parameters include: number of rules, number and type of membership function, changing parameter of membership functions and consequently changing expression of rules according to data distribution in system, type of fuzzy inference system: Mamdani and Sugeno. In investigated system, polymer composition and time are considered as inputs and percent of releasing are consider as output. The obtained results showed that fuzzy parameters could have sufficient effect on modeling discussed system in present study. Designed fuzzy model with Sugeno inference system, Gaussian membership function indicate more potential performance in modeling investigated system. The Obtained mean of squared error (MSE) for test sets in this study proved this fact

In situ forming, controlled drug delivery system, Expert systems, Fuzzy inference systems, MSE