ABSTRACT:
SDU-Haier-AQD (Shandong University-Haier-Appearance Quality Detection) is an image dataset jointly constructed
by Shandong University and Haier, which contains a various of air conditioner external unit image collected
during actual detection process.The Appearance Quality Detection (AQD) dataset is consisted of 10449 images, and
the samples in the dataset are collected on the actual industrial production line of air conditioner. We
selected 11 typical models, marked the typical appearance feature (Brand logo, Connector pipe head and Air
outlet net cover), and divided the feature into 16 labels.
Acknowledgement:
This work is supported by the Major Scientific and Technological Innovation Project of Shandong Province
"Application Demonstration of A New Intelligent Manufacturing Control Platform with Cloud-Edge Collaboration
Mechanism". The research team of Shandong University cooperated withHaier Jiaozhou Air-conditioner
Interconnected Factory build the dataset for appearance quality detection.
Instructions:
The images in the dataset are uniformly in jpg format, and the image size is 2464×2056 or 1232×1028. All the
labels are saved as .xml files. It contains the size of the image, the name and location of the different
labels.
ABSTRACT:
SDU-Haier-ND (Shandong University-Haier-Noise Detection) is a sound dataset jointly constructed by Shandong
University and Haier, which contains the operating sound of the internal air conditioner collected during the
product quality inspection. We collected and marked a batch of quality inspection sounds of air conditioners in
real production environments to form this data set, including normal sound samples and abnormal sound samples.
The data set contains a variety of abnormal sounds, covering a variety of common failures of air conditioners,
such as abnormalities caused by bearings or guide plates.
Acknowledgement:
This work is supported by the Major Scientific and Technological Innovation Project of Shandong Province
"Application Demonstration of A New Intelligent Manufacturing Control Platform with Cloud-Edge Collaboration
Mechanism". The research team of Shandong University cooperated with Haier Jiaozhou Air-conditioner
Interconnected Factory build the dataset for noise detection.
Instructions:
SDU-Haier-ND dataset contains a total of 536 sound samples in wav format, and the sampling frequency of the
sound signal is 48kHz. The sound samples are collected manually, so the sound duration is not the same. The
valid sound signal, that is, the operating sound of the air conditioner, is collected in a confined space called
the sound room to isolate other sounds in the production workshop. The doors of the sound room are open before
and after the air conditioning detection. At this time, the radio equipment in the noise room will collect
various sounds from the production workshop, so the sound samples in this dataset contain some invalid noise.
ABSTRACT:
A dataset asscociated with paper “Learning-based Sparse Data Reconstruction for Compressed Data Aggregation in IoT
Networks” in IEEE Internet of Things Journal. Five different structured sparse models (SSMs) are considered in the
synthesized dataset, including random sparse (Sparse Model A), row sparse (Sparse Model B), row-sparse with embedded
element-sparse (Sparse Model C), row-sparse plus element-sparse (Sparse Model D) and block diagonal sparse (Block Sparse
or group sparse).
If you have any questions or suggestions, do not hesitate to contact the corresponding author: mqzhang@qfnu.edu.cn
Instructions:
Please refer to the paper“Learning-based Sparse Data Reconstruction for Compressed Data Aggregation in IoT Networks”on
IEEE Xplore. https://ieeexplore.ieee.org/abstract/document/9354846
ABSTRACT:
The Prognostic/ Predictive Health Monitoring (PHM) and Fault diagnosis of milling cutters are key issue in the field of machine tool manufacturing. As the "teeth" of CNC machine tools, their health status directly affects the machining efficiency and the quality of products. It has become a hot topic in both academic research and industries, which propose to employ big data and deep learning technology to realize high reliable tool fault diagnosis and predictive maintenance. However, the lack of high-quality yet life-cycle data has become the bottleneck restricting academic research and engineering promotion.
Supported by the Major Scientific and Technological Innovation Project of Shandong Province "R&D and application demonstration of key technologies for new adaptive production systems in the field of machine tool manufacturing", the research team of Shandong University cooperated with Qilu Institute of Technology to jointly setup a multi-sensor based data colloection platform on the machining tools.
At present, the datasets is complete run-to-fault data of 6 end mill.Datasets is open, and it is verified that the tool prediction algorithm of CNC machining tools can be used by anyone.
Instructions:
Please refer to the paper“Quantized Deep Compressed Sensing for Edge-Cloud Collaborative Industrial IoT Networks”on
IEEE Xplore. https://ieeexplore.ieee.org/abstract/document/9869295