AI Summary & Key Takeaways
AI InsightThe theoretical considerations are substantiated by computer experiments and a demonstrative example is presented showing the salient effectiveness of the approach.
📌 3 Key Takeaways
Abstract
multiobjective optimization but also for constraint optimization. Although there are several excellent papers on the penalty function approaches, up till now there is no clear method for the systematic selection of penalty parameters per constraint since the topic is quite elusive. The issues being well-realized, there are several researches addressing these issues to some extent. However, still, the robustness of these methods remains the main issue due to some newly added additional parameters subject to determination. This work endeavours to address this issue and first, it makes a systematic analysis. Following the analysis, it establishes a probabilistic approach as the issue is entirely in the domain of probability. According to the best knowledge of the authors, the approach is unique as to probabilistic treatment of the issue. The approach models the probability density of the random population throughout the generations and based on this, penalty parameters are determined following the probabilistic derivations. The theoretical considerations are substantiated by computer experiments and a demonstrative example is presented showing the salient effectiveness of the approach.