An improved arithmetic optimization algorithm with forced switching mechanism for global optimization problems.

Area of Science:

• Computational Intelligence
• Optimization Algorithms
• Meta-heuristic Methods

Background:

• The Arithmetic Optimization Algorithm (AOA) is a novel meta-heuristic inspired by mathematical operators.
• Original AOA suffers from limited exploration and a tendency to converge to local optima.

Purpose of the Study:

• To enhance the performance of the AOA by improving its exploration capability and ability to escape local optima.
• To introduce an improved AOA (IAOA) incorporating a forced switching mechanism (FSM).

Main Methods:

• The IAOA integrates a forced switching mechanism (FSM) and random math optimizer probability (RMOP).
• RMOP increases population diversity for better global search.
• FSM enables exploratory behavior when agents fail to find better solutions, preventing local optima entrapment.

Main Results:

• The IAOA was tested on 23 classical benchmark functions and 10 CEC2020 test functions.
• Extensive comparisons showed the IAOA outperformed the original AOA and other well-known algorithms on most functions.
• The IAOA proved effective in training multi-layer perceptron (MLP) networks and solving engineering design problems.

Conclusions:

• The proposed IAOA effectively addresses the limitations of the original AOA, particularly local optima convergence.
• The IAOA demonstrates significant improvements in search quality and global exploration.
• The enhanced algorithm shows high potential for solving complex real-world optimization tasks.