A hybrid algorithm of ABC variant and enhanced EGS local search technique for enhanced optimization performance. (September 2018)

• Record Type: Journal Article
• Authors: Sulaiman, Noorazliza; Mohamad-Saleh, Junita; Abro, Abdul Ghani
• Is Part Of: Engineering applications of artificial intelligence. Volume 74(2018)
• Page Start: 10
• Page End: 22
• Publication Date: 2018-09
• Subjects: Artificial bee colony -- Benchmark functions -- Convergence speed -- Economic environmental dispatch -- Global optimum -- Hybrid algorithm; Engineering -- Data processing -- Periodicals; Artificial intelligence -- Periodicals; Expert systems (Computer science) -- Periodicals; Ingénierie -- Informatique -- Périodiques; Intelligence artificielle -- Périodiques; Systèmes experts (Informatique) -- Périodiques; Artificial intelligence; Engineering -- Data processing; Expert systems (Computer science); Periodicals; 620.00285
• DOI: 10.1016/j.engappai.2018.05.002
• Languages: English
• ISSNs: 0952-1976

Abstract

Abstract: A hybrid algorithm for optimizing a complex power system-based problem, economic environmental dispatch (EED) is proposed. The algorithm hybridizes a recently proposed artificial bee colony (ABC) variant referred to as JA-ABC3 and a local search technique, evolutionary gradient search (EGS) which has been enhanced i.e., augmented. The enhanced EGS has been inserted into JA-ABC3's framework and the resulting hybrid algorithm, known as EGSJAABC3 is expected to exhibit robust optimization performance by showing the capability to reach the global optimum in less number of generations. JA-ABC3 which is generated through few modifications towards the standard ABC algorithm is the best candidate as it has exhibited better performance than the standard ABC and other ABC variants. Since JA-ABC3 is a global search algorithm, a local search technique, EGS that has been augmented is selected to be its hybrid partner as it also exhibits better or same performance than its kind. The task of the augmented EGS is to enhance the exploitation capability of the algorithm and thus, guides the solution faster towards the global optimum. In other word, the enhanced EGS is taking part in the exploitation process while JA-ABC3 takes role in exploration and some parts of the exploitation processes. Then, a number of benchmark functions are used to evaluate the robustness of EGSJAABC3 in terms of convergence speed and global optimum achievement. Next, the main significant output of this research is a robust optimization algorithm (i.e., EGSJAABC3) later applied to solve complex real-world problems that is known for their uncertainty. In this paper, EGSJAABC3 is tested to minimize EED on three test generator systems; 6, 10 and 40 units. The acquired outcome on both benchmark functions and EED application demonstrate the robustness of EGSJAABC3 as an optimization algorithm and therefore, provide other researchers and engineers a tool for solving optimization problems. Highlights: A hybrid of ABC variant-enhanced EGS algorithm to solve EED is proposed. JA-ABC3 takes roles exploration and some parts of the exploitation process. The enhanced EGS takes part in the exploitation process. Balanced exploration and exploitation are expected from the hybridization. The results demonstrate the robustness of EGSJAABC3 as an optimization algorithm.