摘要详情

ID / 提交时间

4 / 2025-03-07 18:37:18

标题

Multi-objective Optimization Strategy for the Working Parameters of Chain-tooth Type Residual Film Recycling Machine

关键字

Residual Film Recycling, Neural Network, Genetic Algorithm, Multi-objective Optimization

主题及专题

T14 系统识别与状态估计

状态

全文待审

作者

Penghao Wang / School of Electrical Engineering Xinjiang University

Lirong Xie / School of Electrical Engineering Xinjiang University

Yifan Bian / School of Electrical Engineering Xinjiang University

Zhikang Lin / School of Electrical Engineering Xinjiang University

Long Zhou / School of Electrical Engineering Xinjiang University

Minglei Shi / Xinjiang Bazhou Huifeng Plastics Industry Co., Ltd.

摘要

This study integrates neural networks and genetic algorithms to optimize a chain-tooth residual film recovery machine. Three operational parameters—pickup device angular velocity, separation device angular velocity, and separation angle—were optimized using EDEM-based discrete element simulations and a backpropagation NN model. The non-dominated sorting genetic algorithm (NSGA-II) generated Pareto-optimal solutions, achieving an impurity content rate (ICR) of 24.26% and film recovery rate (FRR) of 83.67%. Strategic parameter selection from the Pareto front balances ICR reduction and FRR retention, enhancing agricultural film recycling efficiency.