The application discloses a rear baffle disc partition
topology optimization method and
system based on multi-feature K means, and the method comprises the following steps: firstly, discretizing a rear baffle disc
design domain and initializing variables; in iteration, generating
physical density by density filtering and Heaviside projection, and solving a
displacement field; in the initial stage, global maximum stress is taken as a constraint for optimization, and after the change rate of the structure
mass is stabilized, a multi-feature K means partition model is switched; the model fuses unit stress, density and
geometric distance features to carry out clustering partition, predicts stress evolution trend based on a transition factor, and dynamically regulates stress constraints of each sub-area by using a sub-area relaxation coefficient and an iterative correction mechanism; and finally, a lightweight topology structure with clear partition and uniform
stress distribution is output. Through the density optimization, feature K means dynamic partition and adaptive
stress regulation technology guided by the transition factor, the application realizes the lightweight
topology design of the rear baffle disc with clear partition, uniform stress and strong manufacturability.