Optimization of rubber mixture production using a validated technological sequence of methods

Abstract

In this study, a combination of Plackett–Burman and Box–Behnken designs is applied to discover the relationships between the components of rubber compounds and technical specifications. Optimization of rubber compound formulation is realized by support vector regression integrated genetic algorithm to minimize compound cost. Twelve components potentially affecting the technical specifications of rubber compound, which are natural rubber, carbon black, white filler, stearic acid, zinc oxide, antiozonant, antioxidant, process oil, curing retarder, curing agent, and accelerator, are screened through Plackett–Burman design to decide the significant variables. Afterwards, four significant parameters, including carbon black, process oil, curing agent, and accelerator are analyzed using Box–Behnken design to minimize the number of experiments while obtaining the correlation between formulation and specifications. Lastly, a support vector regression integrated genetic algorithm is implemented to predict optimum compound formulation at minimum cost.Highlights Optimization of rubber compound to reduce the mixture and curing cost. Combination of Plackett–Burman and Box–Behnken designs. Integration of support vector regression to genetic algorithm. Correlations between the amounts of components and technical specifications.