Genetic algorithms are randomized search algorithms that have been developed in an effort to imitate the mechanics of natural selection and natural genetics. Genetic algorithms operate on string structures, like biological structures, which are evolving in time according to the rule of survival of the fittest by using a randomized yet structured information exchange. Thus, in every generation, a new set of strings is created, using parts of the fittest members of the old set. The main characteristics of a genetic algorithm are as follows:

• The genetic algorithm works with a coding of the parameter set, not the parameters themselves.
• The genetic algorithm initiates its search from a population of points, not a single point.
• The genetic algorithm uses payoff information, not derivatives.
• The genetic algorithm uses probabilistic transition rules, not deterministic ones.

At first, the coding to be used must be defined. Then using a random process, an initial population of strings is created. Next, a set of operators is used to take this initial population to generate successive populations, which hopefully improve with time. The main operators of the genetic algorithms are reproduction, crossover, and mutation.

Reproduction is a process based on the objective function (fitness function) of each string. This objective function identifies how “good” a string is. Thus, strings with higher fitness value have bigger probability of contributing offsprings to the next generation.

Crossover is a process in which members of the last population are mated at random in the mating pool. So, a pair of offsprings is generated, combining elements from two parents (members), which hopefully have improved fitness values. Mutation is the occasional (with small probability) random alteration of the value of a string position. In fact, mutation is a process of random walk through the coded parameter space. Its purpose is to ensure that important information contained within strings may not be lost prematurely.

The implementation of GAs to the problem of optimization of operation and site-scale energy use in production plants is envisaged to be carried along the following lines.

First, using the pinch design method, the problem requirements and constraints will be defined. Next, the region of the pinch will be identified, and the essential matches will be made. Then, the problem will be coded by generating the appropriate strings. These strings will contain the general features and the parameters, which affect the problem. Each string will represent a possible network configuration. For each one of the strings, the objective function will be calculated, that is, the value of energy and utility usage and the number of units need to be used in the represented network configuration, that is, the total capital and operating cost. Initially, a starting population of strings will be created using a random procedure. The three main operators of the genetic algorithms will be performed to improve the value of the objective function, namely, to create network configurations with minimal total capital and operating cost. To the final population of strings/possible networks, advanced techniques will be applied for further improvement.

Regards
Sarah Rose
Managing Editor
International Journal of Swarm Intellgience and Evolutionary Computation