Purpose

For comfort evaluation of underwear pressure, this paper proposes an improved GA algorithm to optimize the weight and threshold of BP neural network, namely PSO-GA-BP neural network prediction model.

Design/methodology/approach

The objective parameters of underwear, body shape data, skin deformation and other data are selected for simulation experiments to predict the objective pressure and subjective evaluation in dynamic and static state. Compared with the prediction results of BP neural network prediction model, GA-BP neural network prediction model and PSO-BP neural network prediction model, the performance of each prediction model is verified.

Findings

The results show that the BP neural network model optimized by PSO-GA algorithm can accelerate the convergence speed of the neural network and improve the prediction accuracy of underwear pressure.

Originality/value

PSO-GA-BP model provides data support for underwear design, production and processing and has guiding significance for consumers to choose underwear.

中文翻译：

---

基于改进GA-BP算法的内衣压力预测研究

目的

针对内衣压力的舒适度评价，本文提出了一种优化BP神经网络权重和阈值的改进GA算法，即PSO-GA-BP神经网络预测模型。

设计/方法/方法

选取内衣客观参数、体型数据、皮肤变形等数据进行仿真实验，预测动态和静态下的客观压力和主观评价。对比BP神经网络预测模型、GA-BP神经网络预测模型和PSO-BP神经网络预测模型的预测结果，验证了各预测模型的性能。

发现

结果表明，采用PSO-GA算法优化的BP神经网络模型能够加快神经网络的收敛速度，提高内衣压力的预测精度。

原创性/价值

PSO-GA-BP模型为内衣设计、生产、加工提供数据支持，对消费者选择内衣具有指导意义。