Most of the optimization problems in practice are multi-objective optimization problems. Usually, the problem of multi-objective optimization lies in the contradiction between each subsystem, and the improvement of one objective usually may cause the degradation of the performance of another or several sub-objectives, which means that it is impossible to make multiple objectives reach the optimal solution at the same time, but only to coordinate and compromise between them so that each subset is optimized as much as possible. Its solution is not unique, but a set of solutions consisting of many Pareto optimal solutions. The aim of the optimization is to make a choice that minimizes the time the system is not in use to the interventions defined in the maintenance plans and also minimizes the time that the system is down for maintenance to occur, but also integrate the life cycle analysis which contains information for the costs and environmental impacts of the design choices.

In this section, we optimize four systems using the nsga2 algorithm. The energy use, CO2, NOx, and SO2 emissions and their resulting cost indicators are compared to the different intervention scenarios and the optimal solution is selected.

We constructed the following functions:

The input dimensions are 11 and they are the intervention items：

CFRP = x[1], M = x[2], CP = x[3], RPA = x[4], RPC = x[5], MAP = x[6], AO = x[7], MPR = x[8], PR = x[9], DR = x[10] , RC = x[11].

The output dimensions are 7 and there are various metrics we consider：”duration”, “interv.dist”, “energy”, “co2″, “nox”, “so2″, “cost”.