The invention relates to the field of complex electromechanical
system health management, in particular to a complex electromechanical
system maintenance opportunity decision-making method under overload operation. The method comprises the following steps: constructing a performance evaluation model based on a belief rule base, defining key feature indexes, overload environment attributes and reference levels thereof, rule weights and attribute weights, and presetting consequent confidence distribution for each rule;
monitoring data are collected, the matching degree of the key feature indexes in each rule is calculated, meanwhile, an overload cumulative index is constructed, and the matching degree of overload environment attributes is calculated; calculating a rule activation weight based on the two types of matching degrees, multiplying consequent confidence distribution by the activation weight, fusing by adopting an
evidence reasoning algorithm to obtain confidence distribution of a
system performance state, and further calculating an output utility value; and determining an
optimal maintenance opportunity based on a preset mapping relation. According to the method, the overload environment attribute is introduced into the activation weight calculation, so that the accuracy of
health assessment under the overload working condition is improved.