200 / 2023-10-20 22:20:11

An Adaptive Real Coded Population-Based Incremental Learning Algorithm for Design Optimizations in Continuous Space

Adaptive updating, evolutionary algorithm, inverse problem, population based incremental learning (PBIL).

6.AI-driven technology > (2)、Intelligent sensing and data analytics for smart manufacturing

Evolutionary Algorithm (EA) has been recognized as the standard and paradigm for solving multimodal functions in continuous spaces. However, the complexity of its genetic operations has prevented its wide applications in engineering design problems. In this regard, increasing efforts have been devoted to EAs based on Probabilistic Models (EAPM) to overcome the aforementioned disruptive effects of available EAs. The Population-Based Incremental Learning (PBIL) algorithm is an EAPM. However, lukewarm efforts have been devoted to PBILs, especially to the real coded PBILs, in computational electromagnetics. In this regard, an adaptive real coded PBIL is introduced. In the proposed PBIL algorithm, a probability matrix is proposed to randomly generate a population, and the updating mechanism for this probability matrix is proposed. Moreover, the interval used in constructing this probability matrix is adaptively updated to strike a balance between convergence performance and solution quality. The proposed PBIL algorithm is finally numerically experimented on a case study with promising numerical results.