Application of a multi objective multi-leader particle swarm optimization algorithm on NLP and MINLP problems

• Presents a constraint handling method independent of penalty and scaling factors.
• Uses an efficient method for handling multiple objectives using PSO.
• Utilizes a modified method for handling discrete variables using PSO.
• Shows the proposed method as a powerful and timely efficient optimization method.
• Applies the proposed algorithm on a variety of chemical engineering optimization problems.

This paper presents a modified particle swarm optimization algorithm for handling a variety of single and multi objective mixed integer nonlinear optimization problems that have equality and inequality constraints. An efficient multi-objective multi leader particle swarm optimization algorithm is used to handle the extra objective imposed by a novel constraint handling method. In addition, a modified method of handling binary variables is used and the algorithm is adapted to update discrete variables independent from continuous variables using these methods. The algorithm was applied on several well known test problems in the field of chemical engineering including the William Otto process. The results proved the applicability and the efficiency of this method for handling single and multi objective optimization problems in mixed integer and nonlinear decision spaces arising in the field of chemical engineering.