An aircraft constrained optimization method based on filter technology and a rectangle division algorithm

Abstract

The invention relates to an aircraft constrained optimization method based on filter technology and a rectangle division algorithm and belongs to the optimization design field in aircraft design. The method comprises the steps of normalizing design space to n-dimensional hypercubes and putting the central points of the n-dimensional hypercubes in a filter; sub-dividing the design space continuously by using the segmentation strategy of the rectangle division algorithm in an iterative process, and obtaining target function values with the central point of each hypercube as a sample point; with the sample points obtained by segmenting the hypercubes as design points, updating elements in the filter according to the dominating thought; judging whether the optimization process is converged, ending the optimization if yes, and, if no, selecting to-be-segmented hypercubes from the filter according to certain conditions to obtain new sample points through segmentation and updating the filter further. The method can obtain the feasible optimal solutions through less tests, thereby saving the computing cost and increasing the design efficiency. The method is great in universality.

Description

technical field [0001] The invention relates to an aircraft constraint optimization method based on filter technology and subdivision rectangle algorithm, which belongs to the field of optimization design in aircraft design. Background technique [0002] Optimum design techniques are increasingly widely used in modern aircraft design to improve system performance and reduce design costs. In practical engineering problems, most of the problems have constraints, that is, optimization problems under constraints. Therefore, it is of great practical significance to study the solution method of the constrained optimization problem for carrying out related design work. In order to solve the constraint problem in the design, the methods of dealing with the constraint conditions commonly used at home and abroad include: penalty function method, sequential quadratic programming and feasible direction method. The latter two methods need to calculate the gradient value of the function...

AI Technical Summary

Problems solved by technology

However, in the actual application combined with the DIRECT or PSO algorithm, it is found that the penalty factor and its update strategy in this method have a greater impact on the optimal design results, and generally require multiple tests to obtain empirical values ​​for specific problems

At present, design simulation analysis in aircraft design often uses high-precision analysis models (such as finite element analysis in structural design, computational fluid dynamics in aerodynamic analysis, etc.), and the design time and cost caused by the repetition in this design are immeasurable of

Images