Computing resources assignment in RTDS simulators with subgraph isomorphism based on genetic algorithm
Qingqi Zhong, Zhigang Wu, Lingxue Lin, Yao Zhang, Jianshe Zhang · 2011
To construct the largest power simulation model with real time digital simulators (RTDS), assignment of the computing resource is an important issue to make the model feasible. For automatic computing resource assignments, the interesting idea of subgraph isomorphism is proposed. Computing units (racks) are connected with communication wires, whereas the power grid is connected with electrical lines. The two systems can be taken as two graphs with physical correspondence. To assign the computing resources is to find a relationship of subgraph isomorphism between two graphs. Genetic algorithm is applied to solve the problem. According to the definition of subgraph isomorphism, the object function is given as the number of ones in the expended incidence matrix defined here. By selection, crossover and mutation in the GA, the problem can be solved and the assignment can also be achieved directly. There are many feasible assignments received.