Optimization and Control of Batch Reactor by Evolutionary Algorithms

This work deals with using a method of artificial intelligence, namely the evolutionary algorithm SOMA, for static optimization of a chemical batch reactor, for the purpose to improve its behavior in uncontrolled state and predictive control. The importance of this problem is increasing with growing demand for special products made in batch reactors. The optimizations has been performed in several ways, each one for another set of reactor parameters or another cost function. The optimization gave the best solution, which improved the performance of the reactor. All the important results of each optimization are discussed continuously and at the end of the paper too. The optimized reactor was used in a simulation with predictive control by the evolutionary algorithms and excellent results were achieved.