Takeo Okazaki

Evolutionary Computing of Petri Net Structure for Cyclic Job Shop Scheduling

This paper considers Cyclic Job-Shop Scheduling Problems (CJSSP) extended from the Job-Shop Scheduling Problem (JSSP). We propose an evolutionary computing method to solve the problem approximately by generating the Petri net structure for scheduling. The crossover proposed in this paper employs structural analysis of Petri net model, that is, the crossover improves the cycle time by breaking the bottle-neck circuit obtained by solving a linear programming problem. Experimental evaluation shows the effectiveness of our approach.

A Two-Phase Metaheuristic for Farm Workscheduling

This paper proposes a two-phase metaheuristic approach to planning daily farm work for agriculture production corporations. The two-phase metaheuristic contains the optimization of resources assignment and searching schedule based on Genetic Algorithm and hybrid Petri nets model. In the experiment, the effect on optimizing the resource assignment and priority list, initializing population of GA with sorted chromosomes by waiting time, inheriting priority list from tasks in the previous resources assignment enhanced the evolutionary speed and solution quality. The computational experiment revealed high effectiveness for constructing farm work schedule with high ratio of resource utilization. The proposed approach also contributes a referential scheme for combining metaheuristic to solve scheduling problem under constraints.

An evolutionary progressive multiple sequence alignment

This paper proposes an evolutionary tree-base (progressive multiple sequence alignment) method using a genetic algorithm (GA) for solving multiple sequence alignment problems. In our evolutionary tree-base method, chromosomes are represented as guide trees. Two kinds of crossover are proposed for chromosomes of tree structure; subtree selection crossover and tree uniform order crossover. They can generate new chromosomes with inheriting tree structure of parents. The indirect representation of multiple alignments, namely, the guide tree representation of chromosomes, and the proper genetic operations make searching drastically efficient. Experimental results for benchmark problems from BAliBASE and the NCBI database show that the proposed method is superior to SAGA (a well-known GA-base approach, 1996), T- coffee (sensitive progressive method, 2000), MUSCLE (progressive/iterative method, 2004), MAFFT (progressive/iterative method, 2005), and ProbCons (probabilistic/consistency method, 2005) with regard to quality of solutions.