# Damage Quantitative Detection of Curved Composite Laminates Based on Improved Particle Swarm Optimization Algorithm

## TL;DR

This paper introduces a new method using an improved particle swarm optimization algorithm to detect and quantify damage in curved composite laminates with high accuracy.

## Contribution

The novel contribution is an improved particle swarm optimization algorithm with boundary rebound and MS operator for accurate damage detection in curved composites.

## Key findings

- The algorithm achieved less than 0.5% detection error in ideal conditions.
- With 1% and 5% noise, the algorithm still located damage accurately with less than 6% error.
- Experimental tests confirmed the algorithm's effectiveness in real-world scenarios.

## Abstract

In order to solve the problem of damage identification of composite laminates during processing and service, a quantitative damage detection method based on swarm intelligence optimization was proposed for structural damage detection of curved composite laminates. Firstly, the structural damage element was defined by the method of reducing the elastic modulus of the element, and the modal parameters of the numerical model of the laminate under different damage conditions were obtained by analyzing the structural vibration characteristics. Secondly, the objective function was constructed from the vibration data, and the precise location and degree of damage were quantitatively calculated by the swarm intelligence optimization algorithm. In order to prevent the particles from falling into the local optimal, the boundary rebound strategy was used to process the boundary, and the MS operator was introduced to greatly accelerate the convergence speed of the algorithm. The numerical results indicate that without the influence of noise, the algorithm was not affected by the quantity, location or size of the damage and could effectively detect damage in curved fiber-reinforced composites, with the detection error rates being within 0.5%. After adding 1% and 5% noise to the frequency and vibration mode, respectively, the convergence speed of the algorithm slowed down, and the convergence times obviously increased. However, it could still accurately locate the damage, and the calculation error of the damage degree was less than 6%. Finally, the effectiveness of the proposed algorithm was verified through experimental tests.