数据驱动的离散智能车间优化运行支持技术数据集

本文发布两套面向智能制造场景的优化数据集，分别支持离散制造动态调度与连续流程工艺优化研究。离散车间动态重构调度仿真实验数据通过构建可重构制造系统的数字孪生模型，生成包含多维机器配置与任务加工参数的.jsr格式仿真数，涵盖12种任务-机器组合维度，采用正态分布参模拟真实生产波动，支持强化学习算法在个性化定制生产中的调度策略优化。工艺参数网络化优化数据集整合2023-2024年多行业生产线传感器数，集成温度（350℃±2%）、压力（2.5MPa）、速度（1200rpm）等工艺参数，结合遗传算法与LSTM预测模型实现网络化优化。两套数据集均通过多重校验机确保数据质量，采用标准化存储格式，为智能制造研究提供多维度实验基准。This paper releases two optimized datasets for intelligent manufacturing scenarios, which respectively support research on discrete manufacturing dynamic scheduling and continuous process optimization. The simulation experimental data for dynamic reconfigurable scheduling in discrete manufacturing workshops is generated by constructing a digital twin model of a reconfigurable manufacturing system, producing .jsr-format simulation data containing multi-dimensional machine configuration and task processing parameters. Covering 12 task-machine combination dimensions, this dataset uses normal distribution parameters to simulate real production fluctuations, supporting the optimization of scheduling strategies for reinforcement learning algorithms in personalized customized production. The networked process parameter optimization dataset integrates sensor data from production lines across multiple industries from 2023 to 2024, incorporating process parameters such as temperature (350℃±2%), pressure (2.5MPa), and speed (1200rpm). It realizes networked optimization by combining genetic algorithms and LSTM prediction models. Both datasets ensure data quality through multiple verification mechanisms and adopt standardized storage formats, providing multi-dimensional experimental benchmarks for intelligent manufacturing research.

搜集汇总

背景与挑战

背景概述

该数据集包含两套面向智能制造场景的优化数据，分别支持离散制造动态调度与连续流程工艺优化研究。它通过数字孪生模型生成仿真数据，并整合多行业生产线传感器数据，采用标准化存储格式，为相关领域提供多维度实验基准。