Multi-sensor IoT data and AI-driven adaptive control model for library indoor environments
- 1Library, Shandong Jiaotong University, Jinan, Shandong, China
Abstract
Modern library areas require constant environmental adjustment to preserve user comfort and promote learning efficiency. This investigation offers a Multi-Sensor Internet of Things (IoT) Data Source and Artificial Intelligence-Based Adaptation Control Model for Library Indoor Systems, which is intended to transform heterogeneous sensing data into intelligent real-time environmental management decisions. A Multi-Sensor IoT Library Indoor Environment Dataset containing 2,745 time-stamped observations was developed, thereby capturing dynamic indoor environmental variations. Normalization and noise smoothing processes were performed before the analysis of the collected data for improving the quality of the analysis. In order to detect the most useful features of the environment, Principal Component Analysis (PCA) was used to perform dimension reduction and simplify calculations. The proposed approach includes edge sensing, intelligent data processing, and actuator cooperation for implementing an efficient closed-loop environmental management system. To classify indoor environmental conditions, a Colony Predation Algorithm enhanced with Random Adaptive Boosting (CPA-RAdaBoost) was developed, where CPA maximizes feature weights and model hyperparameters, Random Forest (RF) enhances ensemble variety, and Adaptive Boosting strengthens the identification of complicated and border environmental conditions. The adaptive control instructions for HVAC, lighting and ventilation are formulated considering the prevailing environmental conditions to facilitate indoor environment control. Experimental implementation using Python demonstrates the effectiveness of the proposed framework, reaching a MAE of 3.9 and control accuracy of 97.1%, indicating its capacity for intelligent, responsive, and energy-efficient smart library environmental management.
Related
- An investigation of novel population models to control aphid outbreaks via a data driven “adaptive” framework
- Adaptive-optics lattice light-sheet microscopy enables correlative single-molecule imaging of endogenous proteins in native environments
- NeuWalk: A Python Library for Data-Driven Synthesis of Neuron Morphologies Using Random Walks
- The adaptive architecture of tRNA dependencies across physiological tumor environments
- Experimental evolution reveals contrasting adaptive landscapes in lab and field environments
BibTeX
@article{zhang2026multisensor,
title = {{Multi-sensor IoT data and AI-driven adaptive control model for library indoor environments}},
author = {Mengxi Zhang},
journal = {PLOS One},
year = {2026},
month = oct,
doi = {10.1371/journal.pone.0361038},
url = {https://doi.org/10.1371/journal.pone.0361038},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Multi-sensor IoT data and AI-driven adaptive control model for library indoor environments
AU - Mengxi Zhang
PY - 2026
DA - 2026/10/27
JO - PLOS One
DO - 10.1371/journal.pone.0361038
UR - https://doi.org/10.1371/journal.pone.0361038
AB - Modern library areas require constant environmental adjustment to preserve user comfort and promote learning efficiency. This investigation offers a Multi-Sensor Internet of Things (IoT) Data Source and Artificial Intelligence-Based Adaptation Control Model for Library Indoor Systems, which is intended to transform heterogeneous sensing data into intelligent real-time environmental management decisions. A Multi-Sensor IoT Library Indoor Environment Dataset containing 2,745 time-stamped observations was developed, thereby capturing dynamic indoor environmental variations. Normalization and noise smoothing processes were performed before the analysis of the collected data for improving the quality of the analysis. In order to detect the most useful features of the environment, Principal Component Analysis (PCA) was used to perform dimension reduction and simplify calculations. The proposed approach includes edge sensing, intelligent data processing, and actuator cooperation for implementing an efficient closed-loop environmental management system. To classify indoor environmental conditions, a Colony Predation Algorithm enhanced with Random Adaptive Boosting (CPA-RAdaBoost) was developed, where CPA maximizes feature weights and model hyperparameters, Random Forest (RF) enhances ensemble variety, and Adaptive Boosting strengthens the identification of complicated and border environmental conditions. The adaptive control instructions for HVAC, lighting and ventilation are formulated considering the prevailing environmental conditions to facilitate indoor environment control. Experimental implementation using Python demonstrates the effectiveness of the proposed framework, reaching a MAE of 3.9 and control accuracy of 97.1%, indicating its capacity for intelligent, responsive, and energy-efficient smart library environmental management.
SN - 1932-6203
ER -