STAIRCASE USAGE CHARACTERISTICS AND MODEL CONSISTENCY BASED ON MULTI-MODEL FUSION
Keywords:
Kernel density estimation, Archard's wear law, Usage frequency, Walking preference, Model consistencyAbstract
To reveal the usage patterns underlying the wear of ancient building stone steps, this paper constructs a multi-model fusion analysis framework. Firstly, pedestrian walking paths are generated using the Monte Carlo method, and a three-dimensional stone step wear simulation model is constructed by combining kernel density estimation (KDE). The normalized KDE probability distribution is integrated into Archard's wear law to form an advanced stone step wear model. Secondly, the usage frequency of the stone steps is analyzed from both the overall and local dimensions. The walking direction preference is determined by geometric center, skewness, centroid, and principal component analysis (PCA). The path structure features are identified using K-means clustering and Gaussian fitting. The research results show that there is no obvious directional preference for stone step usage. The overall distribution is uniform and slightly leans towards the lower part of the Y direction and the right side of the X direction. There are parallel channel characteristics. The model is highly consistent with the existing data in terms of frequency, direction, and path type prediction.References
[1] Song W, Li Y. Deep Learning and Multi-Model Fusion for Analyzing Staircase Wear and Human Traffic Patterns. 2025 7th International Conference on Software Engineering and Computer Science (CSECS). IEEE, 2025: 1-12.
[2] Jayaparvathy R. Modeling of emergency evacuation in high rise buildings considering congestion at stairs based on Markov chains. Physica A: Statistical Mechanics and its Applications, 2024, 633: 129352.
[3] Cheylan I, Favier J, Sagaut P. Immersed boundary conditions for moving objects in turbulent flows with the lattice-Boltzmann method. Physics of Fluids, 2021, 33(9).
[4] Junior B R P, Nagem N F, Filho V C, et al. Optimising Anode Performance in Albras Potlines. Light Metals 2021: 50th Anniversary Edition. Cham: Springer International Publishing, 2021: 708-716.
[5] Geetha M, Vijayabhasker R, Seetharaman S. Computerised Gate Firing Control for 17-Level MLI using Staircase PWM. Computer Systems Science & Engineering, 2023, 44(1).
[6] Song W, Li Y. Deep Learning and Multi-Model Fusion for Analyzing Staircase Wear and Human Traffic Patterns. 2025 7th International Conference on Software Engineering and Computer Science (CSECS). IEEE, 2025: 1-12.
[7] Mei R, Xiang C, Hu Y. Research on Stair Age and Usage Patterns Based on Linear Regression and Maximum Likelihood Estimation. 2025 International Conference on Digital Analysis and Processing, Intelligent Computation (DAPIC). IEEE, 2025: 857-863.
[8] Rapp L, Schmalen L. Error-and-erasure decoding of product and staircase codes. IEEE Transactions on Communications, 2021, 70(1): 32-44.
[9] Das S, Mukherjee S. Navigating cloud security risks, threats, and solutions for seamless business logistics. Emerging technologies and security in cloud computing. IGI Global Scientific Publishing, 2024: 252-275.
[10] Jayawardana J T H, Dilshan H V V, Wijethilaka R, et al. Train a robot to climb staircase using vision-base system. 2022 IEEE 10th Region 10 Humanitarian Technology Conference (R10-HTC). IEEE, 2022: 294-299.