Context awareness by unit-type evolutionary Petri net for team medical care support

The authors propose evolutionary Petri net consisting of units in order to obtain a structural model of team medical care automatically. The number of medical staffs is not sufficient in aged societies, so the work efficiency must be improved. The authors conceive that a Petri net can be used to express cooperative operation; therefore, the structure of a Petri net might assist in understanding the whole team medical care in hospitals and facilities. Evolutionary Petri net is a method to create the best structure of a Petri net that expresses a target workflow well and evolves the genes of individuals correspond to Petri nets. In the previous work, the authors represented each gene as a combination of nodes. However, the representation frequently generated worthless individuals that do not have a Petri net structure. In this work, the authors represent each gene as a combination of units that consist of two types of nodes. Additionally, the authors propose a new method to separately calculate the fitness value on the Petri net structure and that on the unit structures. The authors verified that the total fitness value of the best individual quickly reached the best value by applying genetic operators along with the balance of the two fitness values.