Optimization for a New XY Positioning Mechanism by Artificial Neural Network-Based Metaheuristic Algorithms

Area of Science:

• Mechanical Engineering
• Optimization Algorithms
• Computational Mechanics

Background:

• XY positioning mechanisms are critical in precision engineering.
• Existing optimization methods face challenges in balancing multiple performance objectives.
• Accurate modeling and optimization are essential for compliant mechanism design.

Purpose of the Study:

• To develop and evaluate a novel hybrid optimization method for XY positioning mechanisms.
• To formulate fitness functions and constraints using artificial neural networks (ANN) and particle swarm optimization (PSO).
• To compare the performance of the proposed PSO-GWO and multiobjective optimization genetic algorithm (MOGA) against traditional GWO.

Main Methods:

• Hybridization of artificial neural network (ANN) with particle swarm optimization (PSO) for function formulation.
• Development of a PSO-GWO hybrid algorithm for single-objective optimization.
• Application of multiobjective optimization genetic algorithm (MOGA) for multi-objective optimization scenarios.
• Numerical and experimental validation of the proposed methods.

Main Results:

• The PSO-based ANN method effectively formulates fitness functions.
• PSO-GWO demonstrated superior performance in single-objective scenarios, improving stroke, reducing stress, and increasing the safety factor compared to GWO.
• MOGA provided optimal results for multi-objective optimization, yielding a stroke of 1741.3 μm and a safety factor of 1.8929.
• Prediction results showed good agreement with numerical and experimental verifications.

Conclusions:

• The proposed hybrid optimization techniques (PSO-GWO and MOGA) are effective for optimizing XY positioning mechanisms.
• The integration of ANN with optimization algorithms enhances the modeling and optimization process.
• The findings are expected to aid in the synthesis and analysis of compliant mechanisms and related engineering designs.