Controller Design for Evaluating Optimal Chemical Ratio in Starch Modification Process Mohan M Madhan*, Vijayachitra S** *Assistant Professor, Department of EIE, Kongu Engineering College, Tamil Nadu, India **Professor, Department of EIE, Kongu Engineering College, Tamil Nadu, India Online published on 14 October, 2016. Abstract Native form of starch obtained from Cassava roots is not directly employed due to the property limitations of insolubility in cold water, loss of viscosity and low shear stress resistance. In order to overcome the limitations of native starch and to meet the growing population demand, modified starch manufacturing is essential. Native starch is modified through chemical, physical, enzymatic and genetic methods. Modified starch is used in many applications in textile, pharmacy, plastic, paper, food industries etc. Out of various modified starch production process in an industry, at a time only one particular type of modified starch is manufactured. Currently chemical modification is used in the majority of the starch modification industries which employs manual work for the addition of chemicals with native starch. The chemical modification process requires skilled persons to add precise ratio of chemicals with the native starch according to the various industrial requirements. The improper chemical proportion leads to many health issues especially in food processing industries. In order to reduce this labor-intensive error during manual addition of chemicals with native starch, it is necessary to design a controller to evaluate the proper ratio of chemicals mix with the native starch for the application of modified starch in various industries. The Mamdani Fuzzy Logic Controller (FLC) model and Genetic Algorithm tuned Fuzzy Logic Controller (GFLC) model designed for modified starch processing industry to evaluate the precise chemical ratio required to meet the applications of modified starches in Pharmacy, Paper, Plastic and Food industries. Here four input parameters of native starch (pH, Temperature, Viscosity, Moisture) and Type of industry act as input for the FLC/GFLC and three output parameters (Sodium Hydroxide, Hydrochloric Acid and Malic Anhydride/Octenyl Succinic Anhydride/Phthalic Anhydride/Sodium Hypochlorite) are considered as output from FLC/GFLC. Sodium Hydroxide and Hydrochloric acid employed for pH neutralization, Phthalic Anhydride/Sodium Hypochlorite/Malic Anhydride/Octenyl Succinic Anhydride employed to maintain viscosity. In FLC and GFLC model, centroid method is employed for defuzzification and the chemical ratio obtained from simulated results are compared in terms of maximum output error. The maximum output error from FLC model found to be 18.75%, 50%, 15% and 20.625% for Pharmacy, Paper, Plastic and Food industrial applications respectively. The results shows that GFLC model has improved performance in reducing the maximum error from 50% to 2.5% for Paper Industry applications and FLC model is better than GFLC for modified starch production in Pharmacy, Plastic and Food industrial applications. Top Keywords FLC-Fuzzy Logic Controller, GA-Genetic Algorithm, GFLC-Genetic Algorithm tuned Fuzzy Logic Controller, HCl-Hydrochloric Acid, C4H2O3-Malic Anhydride, C12H18O3-Octenyl Succinic Anhydride, C8H4O3-Phthalic Anhydride, NaOH-Sodium Hydroxide, NaClO-Sodium Hypochlorite, PCSPh-Pregelatinized Cassava Starch Phthalate, CPS-centipoise, L-Low, M-Medium, H-High, VL-Very Low, VH-Very High. Top |