Bio-inspired multi-agent systems for reconfigurable manufacturing systems
Engineering Applications of Artificial Intelligence
| Hi-index | 0.00 |
The purpose of this synthesis is to create an overall picture about the control algorithms based on the evolutionary model. These methods offer a new and modern approach of solving complex problems seen in terms of automation and electrical drives controlled by computer algorithms. The modular approach also allows easy automation procedure: the genetic algorithm receives some parameters, which are processed and then will be selected the best results based on information provided by the evaluation function. This method allows a better response of the system by making changes to the evaluation function. In the case of using on-line genetic algorithms, by changing the evaluation function, results are generated in real time, the only condition is to ensure sufficient computing capacity in order to execute the genetic algorithm (in fact the execution of genetic algorithm and the response from the evaluation function must be in the interval that allows the system to react in real time). Because it takes a large computing capacity for running the program in a very short time, genetic algorithms are not suitable in solving small problems with simple solutions because there will be always solutions that are cheaper and easier to implement. But if are considered complex problems, which don't have a clear solving algorithm, or there are some problems with few known data, the genetic algorithms and evolutionary methods in general are more suitable because they can offer solutions, which are otherwise difficult to obtain.