Analysis of ANFIS and ANN for Building Automated Surrogate Algorithms

While attaining the objective of online optimization of complex chemical processes, the possibility of using the first principle based models is rarely an option, since such models demand large computational time. Surrogate models, which can emulate first principle based models, offer a credible solution to this problem by ensuring faster optimization. Thus, the entire challenge of enabling online optimization of complex models depends on construction of efficient surrogate models. Often, the surrogate building algorithms have certain parameters that are usually fixed based on some heuristic, thereby inviting potential errors in building such surrogate models. This work aims at presenting an elaborate study on the effect of various parameters affecting the predictability of Adaptive Neuro Fuzzy Inference Systems viz. (a) architecture of ANFIS, (b) sample size required by the ANFIS, (c) maximum possible accuracy of prediction, (d) a robust sampling plan. The ANFIS is then utilized as surrogates for a highly nonlinear industrial PVAc process, the optimization of which is then realised nearly 9 times faster than the optimization study using the expensive phenomenological model. A brief study was also conducted on another well-known class of surrogates, Artificial Neural Networks, for modelling of the Electrospinning process