The invention relates to the technical field of automobile air conditioner
system design and optimization, in particular to an automobile air conditioner
HVAC system multi-target design method based on
sensitivity analysis and collaborative optimization. Screening out design variables which are respectively strongly related to the
air volume and the aerodynamic
noise;
kriging approximation models are constructed for the two subjects of
air volume and
noise respectively, single-target optimization is carried out by adopting a multi-island
genetic algorithm, and the optimal solution of each target in a
design space is determined; an improved collaborative optimization strategy framework is constructed, a
system-level optimization target is set as a weighted normalization function of
air volume maximization and
noise minimization, a
relaxation factor is introduced to relax consistency constraint from an equation to an inequality, and minimization of difference of design variables between a
system level and a subject level is taken as a subject-level target. Iterative optimization is carried out, and finally the optimal
balance design of the
centrifugal fan of the
HVAC system is obtained.