Change Serial Number Lenovo Thinkcentral Mt

29/11 0 By bronell

Change Serial Number Lenovo Thinkcentral Mt

Secondly, the NSGA II also has some superior properties such as the handling of not objective functions. For example, the number of the Pareto fronts in the considered design system is smooth increasing, and there is no over or under sizing of the Pareto front number. In addition, the NSGA II has advantages over other developed on the NSGA III in the handling of non objective functions. For example, the variation trend of each domain (i.e., a micro) or each set of data (i.e., a setting) is not noticeable in the initial stage of execution. This is because the NSGA II divides each problem into the sub problem, and then it executes the initial cycle three times. The initial cycle can be terminated on the basis of performance or another criteria if it is needed. The NSGA II can also monitor and control excessive performance loss, and it can restart the execution if the stability of the system is disturbed.

Firstly, NSGA-II is tested to compare with PSO and sequential scheduling. Fig 4 shows that the average execution time of NSGA-II is the shortest when there are five tasks. The PSO performs slightly higher than the sequential scheduling, and Fig 6 shows that the tasks are dispatched to helper with best task processing rate in every scheduling. As the number of tasks increases, the PSO introduces the greatest delay. Fig 7 suggests that Fig 11 and Fig 8 present a similar result, which is that Fig 9 is a merging line. Especially, when there are twenty tasks, the PSO runs far more inefficiency than the sequential scheduling, and Fig 11 indicates that the total execution time of sequential scheduling and NSGA-II are identical because of the running to the end state. In addition, we can also see that when there is less than ten tasks, Fig 12 presents that Fig 3 has no plateau, and that there is an obvious plateau in Fig 6, where it is only found that the optimal number of iteration times of NSGA-II is 15 when the number of tasks and helpers are fixed. As can be seen from Fig 8, when the number of tasks is fixed, less than 15 iterations take almost the time of sequential scheduling. That means the whole system executes within 15 iterations, and running to the end state is no longer necessary.