The invention discloses an
engineering structure multi-objective intelligent optimization method based on discontinuous variables. The method comprises the steps that a multi-objective optimization model is constructed, and the discontinuous variables in the multi-objective optimization model are selected; selecting representative sample points in the discontinuous variable feasible region based on central point composite
test design, and fitting second-order polynomial response surface
model parameters through a least square method to obtain an agent model; a
population is initialized through real number coding, single-point
crossover operation and single-
gene random reset
mutation operation based on discontinuous variables are designed, and a solution set is ensured to be matched with a feasible region of a project; based on an elitism strategy and a fast non-dominated
sorting algorithm, sorting is carried out, a
Pareto optimal solution set is determined, and a balanced optimal structure design scheme of the
Pareto optimal solution set is selected. According to the method,
test design, an agent model and an intelligent
algorithm are integrated, the optimization efficiency is remarkably improved, the precise application problem of discontinuous variables in
engineering is solved, collaborative optimization of performance and economy is achieved, and the method is suitable for lightweight and long-life design of
engineering structures such as bridges and buildings.