Hybrid Approach of Genetic Programming and Quantum-Behaved Particle Swarm Optimization for Modeling and Optimization of Fermentation Processes

Abstract

This paper proposes a novel method for modeling and optimization of fermentation process with a hybrid approach of genetic programming (GP) and quantum-behaved particle swarm optimization (QPSO). In this method, the GP algorithm is first used to model the process, with the parameters of the model selected randomly within a given interval, while the population of models evolves. Then, the parameters of the model obtained by the GP are tuned by the QPSO algorithm in order to increase the fitting accuracy. Finally, the values of the independent variables of the model representing the culture conditions are optimized by the QPSO in order to maximize the dependent variable, which generally represents the yield of the fermentation product. The proposed method is applied to the fermentation process of the hyaluronic acid (HA) production by Streptococcus zooepidemicus. The experimental results show the efficiency of the GP-QPSO approach in the modeling and optimization of this fermentation process.