Determining and Comparing the Optimal Amount of Fat and Balangu Seed Gum in Fresh Yogurt Using Two Methods of Multiple Objective Particle Swarm Algorithm and Response Surface Methodology

In this study, the formulation of fresh low-fat yogurt containing Balangu seed gum was modeled and optimized. In this study, first the independent variables of balangu seed gum (0 to 0.1%) and fat (0 to 2%) were modeled completely randomly in the form of a central composite design and each of the response variables (syneresis, pH, flavor, color, texture, and overall acceptance) were presented in the form of a polynomial regression model as a function of independent variables.The results showed that with increasing balangu seed gum, syneresis and pH of the samples decreased and the flavor, color, texture, and overall acceptance of the samples increased. Also, the increase in fat only decreased the syneresis of the samples. Then, the obtained models were optimized by multi objective particle swarming algorithm and numerical optimization algorithm in response surface methodology, so that minimum syneresis and maximum texture, flavor, color and general acceptance scores were obtained. The results of comparing the mean of three optimal points of the two algorithms showed that in general, in optimizing the formulation of low-fat yogurt containing balangu seed gum, the multi-objective particle swarming algorithm has a better performance than the numerical optimization algorithm in response surface method. The average optimal amount of balangu seed gum and fat in the multi-objective particle swarm algorithm was 0.85 and 1%, respectively, and in the method of optimization algorithm in response surface methodology, was 0.89 and 1.94%.