MINIMUM WEIGHT DESIGN OF FRAMED STRUCTURES USING A GENETIC ALGORITHM CONSIDERING DYNAMIC ANALIYSIS

Abstract

GENETIC ALGORITHMS CAN BE CONSIDERED AS A FAMILY OF COMPUTATIONAL MODELS INSPIRED ON THE PRINCIPALS OF EVOLUTION. THESE ALGORITHMS START FROM A SET OR POPULATION OF CHROMOSOMES IN WHICH EACH CHROMOSOME CONTAINS A POSSIBLE SOLUTION TO THE SPECIFIC PROBLEM. GENETIC OPERATORS, (SELECTION, CROSSOVER AND MUTATION) ARE APPLIED WHICH GENERATE A NEW POPULATION PRESERVING AND IMPROVING THE INFORMATION CONSIDERED IN THE ORIGINAL CHROMOSOMES. HERE THE OPTIMIZATION OF THE CROSS-SECTIONAL AREA OF FRAME STRUCTURES IS CONSIDERED IN ORDER TO OBTAIN THE MINIMUM WEIGHT FOR DYNAMIC ANALYSIS. THE OPTIMIZATION IS OBTAINED THROUGH THE MINIMIZATION OF AN OBJECTIVE FUNCTION, WHICH, FOR THE EXAMPLES CONSIDERED IS WRITTEN USING THE CROSS-SECTIONAL AREA AND THE LENGTH OF THE BARS AND THE SPECI¯C WEIGHT OF THE MATERIAL FROM WHICH THE TOTAL WEIGHT OF THE STRUCTURE CAN BE OBTAINED. THE CROSS-SECTIONAL AREAS ARE THE PROJECT VARIABLES WHICH SHOULD BE WITHIN CERTAIN LIMITS IN ORDER TO OBTAIN REALISTIC SOLUTIONS. RESTRICTIONS ARE APPLIED ON THE NATURAL FREQUENCIES. RESULTS ARE OBTAINED FOR SOME TEST PROBLEMS.