A search algorithm for constrained engineering optimization and tuning the gains of controllers

In this work, the application of an optimization algorithm is investigated to optimize static and dynamic engineering problems. The methodology of the approach is to generate random solutions and find a zone for the initial answer and keep reducing the zones. The generated solution in each loop is independent of the previous answer that creates a powerful method. Simplicity as its main advantage and the interlaced use of intensification and diversification mechanisms-to refine the solution and avoid local minima/maxima-enable the users to apply that for a variety of problems. The proposed approach has been validated by several previously solved examples in structural optimization and scored good results. The method is also employed for dynamic problems in vibration and control. A modification has also been done on the method for high-dimensional test functions (functions with very large search domains) to converge fast to the global minimum or maximum; simulated for several well-known benchmarks successfully. For validation, a number of 9 static and 4 dynamic constrained optimization benchmark applications and 32 benchmark test functions are solved and provided, 45 in total. All the codes of this work are available as supplementary material in the online version of the paper on the journal website.

Automated summary

This research investigates the application of an optimization algorithm in optimizing static and dynamic engineering problems. The algorithm generates random solutions, continuously refines them, and has been validated in structural optimization and dynamic problems such as vibration and control. The method's simplicity, along with its ability to avoid local minima/maxima, enables its application for a variety of problems.