Drawing Graphs, Scheduling, Partitioning, and Path Planning

As stated in the Introduction, it seems that most researchers “modified” their implementations of genetic algorithms either by using non-standard chromosome representation or by designing problem-specific genetic operators (e.g., [104], [294], [48], [56], etc.) to accommodate the problem to be solved, thus building efficient evolution programs. The first two modifications were discussed in detail in the previous two chapters (Chapters 9 and 10) for the transportation problem and the traveling salesman problem, respectively. In this chapter, we have made a somewhat arbitrary selection of a few other evolution programs developed by the author and other researchers, which are based on non-standard chromosome representation and/or problem-specific knowledge operators. We present systems for the graph drawing problem (Section 11.1), scheduling problems (Section 11.2), the timetable problem (Section 11.3), and partitioning problems (Section 11.4). The chapter concludes with a section on path planning problem in mobile robot environment (Section 11.5). The described systems and the results of their applications provide an additional argument to support the evolution programming approach, which promotes creation of data structures together with operators for a particular class of problems.