Hybridization of nonlinear sine cosine and safe experimentation dynamics algorithms for solving control engineering optimization problems
In the rapidly developing landscape of control engineering within electrical and electronics engineering, the study addresses critical challenges posed by the escalating complexity of system designs. As technology continues to advance, the demand for effective methodologies to design, analyse, and s...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2024
|
| Subjects: | |
| Online Access: | http://umpir.ump.edu.my/id/eprint/42486/1/ir.Hybridization%20of%20nonlinear%20sine%20cosine%20and%20safe%20experimentation%20dynamics%20algorithms.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | In the rapidly developing landscape of control engineering within electrical and electronics engineering, the study addresses critical challenges posed by the escalating complexity of system designs. As technology continues to advance, the demand for effective methodologies to design, analyse, and synthesize complex system models becomes vital. In response to this imperative, researchers have increasingly turned to optimization-based approaches, with the Sine Cosine Algorithm (SCA) emerging as a prominent solution. However, existing limitations in terms of convergence accuracy and local optima stagnation have prompted the introduction of innovative improvements in this study. The Nonlinear Sine Cosine Algorithm (NSCA) is introduced as a key extension, incorporating a versatile nonlinear decreasing gain into the transition parameter mechanism. This enhancement offers a tailored balance between exploration and exploitation, aligning with the specific requirements of complex electrical and electronics engineering problems. Moreover, the Nonlinear Sine Cosine Algorithm-Safe Experimentation Dynamic (NSCA-SED) introduces a hybridization of multi-agent and single-agent algorithms, presenting a dynamic approach with random perturbation to navigate search trajectories effectively, and release design parameters that might be trapped in local optima. The empirical assessment of these proposed methods encompasses a diverse set of 23 benchmark functions, demonstrating their efficacy comparable to well-established metaheuristic algorithms such as as the Grey Wolf Optimizer (GWO), Multi-Verse Optimization (MVO), Sine Cosine Algorithm (SCA), Ant Lion Optimizer (ALO), Moth-Flame Optimization Algorithm (MFO), and Grasshopper Optimization Algorithm (GOA). The applications extend beyond established optimization problems, addressing contemporary experiments in control engineering, including Model Order Reduction, Nonlinear System Identification, and Data-driven Control. Simulation results underscore the robustness and superiority of the NSCA and NSCA-SED algorithms in these contexts, showcasing improvements ranging from 13.97% to 97.17% for Model Order Reduction, 17.76% to 99.37% for Nonlinear System Identification, and 84.51% to 89.47% for Data-driven Control when compared to the standard SCA. In summary, this study not only contributes advancements to optimization algorithms but also directly addresses and enhances methodologies in electrical and electronics engineering. By overcoming the limitations of existing approaches, the NSCA and NSCA-SED algorithms stand as valuable tools in the collection of control engineers, facilitating the design and optimization of complex systems in contemporary electrical and electronics applications. |
|---|