{"title":"Gradual Innovative Transitional Solutions Improving Current to a Desired Target: Innovation Path","authors":"Ahmer Khan;Kalyanmoy Deb","doi":"10.1109/TEVC.2025.3552685","DOIUrl":null,"url":null,"abstract":"In practice, there is often a need to update the currently implemented (CI) solution to achieve better performance goals catering to new demands or adoption of new technologies. However, the new optimal solution, found by reoptimizing the problem, may be quite different from the CI solution implicating large costs, major changes, and laborious efforts, causing an apathy for its adoption. For such scenarios, we propose a concept of an “innovation path” (IP), containing a sequence of transitional solutions from the existing to the new target solution with gradual and controlled change from one to the next. To discover such intermediate solutions of the IP, we propose a bi-objective formulation with dynamic step-constraints as an IP problem (IPP), such that a finite set of Pareto-optimal solutions of the resulting IPP become the desired intermediate IP solutions. Due to required gradual discovery of IP solutions, the IP-seeking task happens to be a nontrivial task. We demonstrate the working of the proposed approach on a number of single-, two-objective, and many-objective test and engineering problems. This article concludes with a number of extensions of this study, but the results of this study clearly indicate the usefulness of the proposed approach to other practical problems.","PeriodicalId":13206,"journal":{"name":"IEEE Transactions on Evolutionary Computation","volume":"30 1","pages":"378-392"},"PeriodicalIF":11.7000,"publicationDate":"2025-03-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10930915","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"IEEE Transactions on Evolutionary Computation","FirstCategoryId":"94","ListUrlMain":"https://ieeexplore.ieee.org/document/10930915/","RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE","Score":null,"Total":0}
引用次数: 0
Abstract
In practice, there is often a need to update the currently implemented (CI) solution to achieve better performance goals catering to new demands or adoption of new technologies. However, the new optimal solution, found by reoptimizing the problem, may be quite different from the CI solution implicating large costs, major changes, and laborious efforts, causing an apathy for its adoption. For such scenarios, we propose a concept of an “innovation path” (IP), containing a sequence of transitional solutions from the existing to the new target solution with gradual and controlled change from one to the next. To discover such intermediate solutions of the IP, we propose a bi-objective formulation with dynamic step-constraints as an IP problem (IPP), such that a finite set of Pareto-optimal solutions of the resulting IPP become the desired intermediate IP solutions. Due to required gradual discovery of IP solutions, the IP-seeking task happens to be a nontrivial task. We demonstrate the working of the proposed approach on a number of single-, two-objective, and many-objective test and engineering problems. This article concludes with a number of extensions of this study, but the results of this study clearly indicate the usefulness of the proposed approach to other practical problems.
期刊介绍:
The IEEE Transactions on Evolutionary Computation is published by the IEEE Computational Intelligence Society on behalf of 13 societies: Circuits and Systems; Computer; Control Systems; Engineering in Medicine and Biology; Industrial Electronics; Industry Applications; Lasers and Electro-Optics; Oceanic Engineering; Power Engineering; Robotics and Automation; Signal Processing; Social Implications of Technology; and Systems, Man, and Cybernetics. The journal publishes original papers in evolutionary computation and related areas such as nature-inspired algorithms, population-based methods, optimization, and hybrid systems. It welcomes both purely theoretical papers and application papers that provide general insights into these areas of computation.