Implementation of Covariance Function to Improve Ant Colony Algorithm for Common Chemical Engineering Optimization

Orn-anong Winyutrakoon

Saharat Rattanapairom

Pichita Petpraphan

Thongchai Rohitadisha SrinophakunEmail

Department of Chemical Engineering, Faculty of Engineering, Kasetsart University, Bangkok 10900, Thailand.

Abstract

An improved real-value ant colony optimization algorithm (IACOR) focused on a weight associated with the solution by applying the covariance function to dominate the probability of selection. The performance of the developed algorithm was validated using 9 common engineering benchmarks, followed by 19 classical benchmarks. Furthermore, IACOR was introduced to typical chemical process simulations and developed to work seamlessly with a process simulator (ASPEN), as illustrated in 2 additional examples of distillation column and heat exchanger network. IACOR produced better optima and decreased the computational time with a lower number of iterations to solve complicated problems compared to the conventional method. The highest error compared to some available exact solutions are not over 1 % at low dimension of decision variables.

Implementation of Covariance Function to Improve Ant Colony Algorithm for Common Chemical Engineering Optimization