| description abstract | As an open and commonly employed file format in building information modeling (BIM), industry foundation classes (IFC) have been supported by most BIM software platforms. However, IFC files exported from IFC-supported software usually include numerous duplicate data instances, which limit its application in IFC-based data storage, exchange, transmission, and others. To contend with this issue and optimize IFC files, this paper proposes an automatic content compression algorithm based on reference relationships among data instances (ACC4IFC). The aim is to remove the redundant data instances efficiently. The presented algorithm was implemented by multiple recursive traversal processes based on the IFC model graph. The compression procedure was lossless, and only redundant information was eliminated from the original IFC file. The algorithm was tested using multiple IFC files of various size. The results confirm that the proposed ACC4IFC algorithm is superior to the previous approaches. When compared with the Solibri IFC Optimizer, the ACC4IFC algorithm is shown to remove more redundancies in IFC files. In contrast to the IFCCompressor, the ACC4IFC algorithm is shown to save 66.293% consumption time on average. | |
