Optimization of multistage vapour compression systems using genetic algorithms. Part 2: Application of genetic algorithm and results

Genetic algorithms involve the coding of a solution into a binary string in the same manner that DNA is a biological coding. A population of binary strings are randomly created, evaluated, allowed to mate and are mutated to form a new generation of strings. There is a mating preference given to those strings which rate the highest to simulate the survival-of-the-fittest theory that exists in nature. This process of evaluation, mating and mutation is repeated until some termination criteria are met. A computer code was written in Visual C++ to simulate the vapour compression systems and perpetuate the genetic algorithm. The genetic algorithm functioned adequately enough to provide general trends but it did not find a universal optimum. After numerous runs, the code produced data that suggest that systems which employ intercooler/flash tanks and operate at lower evaporating temperatures have a higher multistage effectiveness.