Because the structure is increasingly complex, and the
investment is getting higher and higher, it is unrealistic to carry
out several reliability experiments. Simultaneously, due to the
application of preventive maintenance, the fault data gradually
decreases. As a result, the above reasons bring about the lack of
CBM data. Furthermore, the engineering background of
electrical and mechanical equipment is relatively complex. The
evaluation needs large-scale model, much more data and
variables that the traditional mathematical methods are unable
to achieve. Therefore, it establishes the genetic algorithm
model based on support vector to solve the optimization model.
The figure 5 shows the detailed optimization process of
maintenance program in the running phase.