Optimization of industrial CSTR for vinyl acetate polymerization using novel shuffled frog leaping based hybrid algorithms and dynamic modeling

Industrial scale continuous stirred tank reactor (CSTR) for production and grade transitions of poly vinyl acetate (PVAc) at the different reactor sizes was investigated. Such reactor is known to show oscillatory behavior and to have periodic limit points, particularly at high molecular weights. Four efficient novel hybrid optimization methods which use variable population size genetic algorithm (VPGA), bacterial optimization algorithm (BO) and shuffled frog leaping method (SFL) were introduced for this kind of reactors. These algorithms can reliably find dynamically stable points with desired conditions for PVAc production. The dynamic modeling combined with the new hybrid optimization methods were used to obtain the optimal operating conditions which increased the outlet flow rate by 17%.

► We investigated the level of molecular weight of PVAc on the operation of the CSTR at the different reactor sizes. ► The various optimization techniques including BO, VPGA and SFL in association with dynamic modeling were examined to analyze the system. ► The novel hybrid methods based on SFL algorithm were developed to achieve the best operating conditions of the CSTR for producing the desired product.