Abstract:Abstract With the rapid development of our country’s economy and industry, atmospheric environmental issues have gradually attracted more and more people’s attention. In recent years, the vigorous development of the Internet of Things technology has made it possible to “Internet of Everything”. At present, the common traditional atmospheric environment monitoring systems on the market generally have shortcomings such as single function, large system measurement error, inability to collect data for a long time, large system power consumption, and inability to view data after being far away from the device. As Bayesian network theory is more and more applied to the modeling of monitoring events, more and more learning optimization methods about Bayesian network have been proposed and become a research hotspot. The purpose of this paper is to study the design of atmospheric environmental health monitoring system based on spatial Bayesian network. This paper studies the intelligent anomaly detection technology based on Bayesian network, the purpose is to use the intelligent anomaly detection technology to predict unknown behaviors that affect the health of the atmospheric environment, improve the learning ability of the detection system, and strengthen the active defense capability of the atmospheric environmental health monitoring system; Secondly, this article proposes an improved spatial Bayesian network model based on the shortcomings of the spatial Bayesian network. Finally, in view of some of the shortcomings of the detection system, this paper proposes a system model that combines anomaly detection with other detection mechanisms, which can significantly improve the accuracy of detection and reduce the false alarm rate of the atmospheric environmental health monitoring system. This paper aims at the above shortcomings and at the same time integrates the current situation of the atmospheric environment monitoring system, analyzes the possible trends of the atmospheric environment monitoring system, and combines the advantages of the spatial Bayesian network to address temperature, humidity, carbon monoxide concentration, and suspended particle concentration. For the main atmospheric environment indicators, a multi-functional, low measurement error, low power consumption, and low-cost atmospheric environment monitoring system is designed. Experimental research results show that: in terms of concentration, nitric oxide and nitrogen dioxide are the highest, at 58% and 68%, respectively, and they are also the largest in the air, accounting for a total of 70%. All in all, the monitoring system needs to monitor even more types of gases listed above, so that the feedback of the atmospheric environment health status is more accurate.

Modeling and Monitoring of a Multivariate Spatio-Temporal Network System

Distributed Self-Monitoring Sensor Networks Via Markov Switching Dynamic Linear Models

Traffic Sensor Health Monitoring Using Spatiotemporal Graphical Modeling

Process Monitoring Based On Multivariate Causality Analysis And Probability Inference

AutoST: Towards the Universal Modeling of Spatio-temporal Sequences

Exploring spatio-temporal correlation and complexity of safety monitoring data by complex networks

Anomaly detection using spatial and temporal information in multivariate time series

Traffic Dynamics Exploration and Incident Detection Using Spatiotemporal Graphical Modeling

Modelling and monitoring multi-relational networks with ordinal information

Damage Detection of Bridge Network with Spatiotemporal Pattern Network

Spatiotemporal Polynomial Graph Neural Network for Anomaly Detection of Complex Systems

Computer Network Security Analysis Modeling Based on Spatio-temporal Characteristics and Deep Learning Algorithm

Monitoring industrial control systems via spatio-temporal graph neural networks

Dynamic Field Monitoring Based on Multitask Learning in Sensor Networks

A Space-Time Network-Based Modeling Framework for Dynamic Unmanned Aerial Vehicle Routing in Traffic Incident Monitoring Applications

A Multi-Sensor Mapping Bi-LSTM Model of Bridge Monitoring Data Based on Spatial-Temporal Attention Mechanism

Traffic System Anomaly Detection Using Spatiotemporal Pattern Networks

Spatio-temporal Modeling for Massive and Sporadic Access

A Novel Stochastic Modeling Method For Network Security Situational Awareness

Design of Atmospheric Environmental Health Monitoring System Based on Spatial Bayesian Network

Multi-Scale Traffic Aware Cybersecurity Situational Awareness Online Model for Intelligent Power Substation Communication Network