Full Constraint Parallel Optimization Algorithm for Structural Optimization of Product Family

This paper discusses the problem of structural optimization of product family, and a full constraint parallel optimization (FCPO) algorithm is devised for this problem. The FCPO algorithm decomposes the product family optimization problem into sub-problems, each sub-problem only optimizes the variables of common parts and one product's unique variables, the variables of common parts are kept consistent during the process of optimizing, and the variables of unique parts in other products are fixed as constants, which are equal to the optimal results of the last iteration. The objective of each sub-problem is to minimize all products' weight, and each sub-problem contains the constraints of all products. This paper optimizes a truss family with different algorithms, the result shows that the FCPO algorithm is the most effective algorithm.