TRANSYT-7F

Genetic Algorithm Optimization

Genetic algorithm optimization is a significant improvement over the traditional hill-climb optimization technique that has been employed by TRANSYT-7F for many years.  The genetic algorithm has the ability to avoid becoming trapped in a "local optimum" solution, and is designed to locate the "global optimum" solution. TRANSYT-7F features genetic algorithm optimization of cycle length, phasing sequence, splits, and offsets.


Genetic Algorithm Output 


For years, engineers have known that they could obtain better timing plans by tweaking the so-called "optimal result" recommended by a computer program.  This is a byproduct of the hill-climb optimization process, where most timing plan candidates are not examined, in an effort to save time.  The global optimum solution is often skipped over during the hill-climb optimization process.

Now, with genetic algorithm optimization, the engineer may have a more difficult time in coming up with a better solution than the computer program.  The genetic algorithm does not examine every single timing plan candidate either, but is a random guided search, capable of intelligently tracking down the global optimum solution.  As with the human race, the weakest candidates are eliminated from the gene pool, and each successive generation of individuals contains stronger and stronger characteristics.  Itís survival of the fittest, and the unique processes of crossover and mutation conspire to keep the species as strong as possible.

Although it produces the best timing plans, a practical disadvantage of the genetic algorithm involves longer running times on the computer.  Fortunately, this disadvantage continues to be minimized by the ever-increasing processing speeds of today's computers.  In addition, the TRANSYT-7F documentation offers practical suggestions for reducing optimization running times.

Take the "G.A. Challenge":  After recording hundreds of test cases, no input file has ever been found for which hill-climb optimization produces a better timing plan than genetic algorithm optimization.  If you encounter a TRANSYT input file for which hill-climb optimization appears to produce a better result, submit it to McTrans technical support by e-mail.  We will attempt to return it genetic-optimized with the best solution.