Method for tracing and quantitatively evaluating unorganized emission

By combining a multi-agent system and a sensor network with a diffusion model, the challenges of monitoring and quantifying fugitive emissions have been solved, enabling high-precision source tracing and environmental accountability, and supporting closed-loop environmental management.

CN122222787APending Publication Date: 2026-06-16CHONGQING UNIV OF POSTS & TELECOMM
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
CHONGQING UNIV OF POSTS & TELECOMM
Filing Date
2026-04-23
Publication Date
2026-06-16

AI Technical Summary

Technical Problem

Existing technologies are insufficient to accurately monitor and quantify fugitive emissions in industrial production; traditional methods lack precision, and high-cost technologies are difficult to implement routine monitoring.

Method used

A multi-agent system is used to process heterogeneous data sources. By combining random forest and particle swarm optimization algorithms, data mining and localization are performed through sensor networks and diffusion models to generate contour maps of pollutant concentration contributions, providing a basis for environmental liability determination.

Benefits of technology

It enables high-precision quantification and source tracing of fugitive emissions, provides scientific environmental accountability reports, and supports closed-loop management of environmental management.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122222787A_ABST
    Figure CN122222787A_ABST
Patent Text Reader

Abstract

The present application relates to a kind of unorganized emission traceability and quantitative evaluation method, belong to atmospheric environment monitoring and pollution source analysis technical field.It includes: collecting the historical data of enterprises in the industry and preprocessing, constructs standard feature matrix in the same industry;Based on standard feature matrix in the same industry and operating multidimensional feature vector matrix, unsupervised anomaly mining is carried out using random forest algorithm, and marked emission abnormal enterprises;For the marked enterprise, sensor network is deployed to obtain real-time field data, and random forest regression model is used to correct sensor compensation parameters;Sensor data is input into Gaussian diffusion model, combined with particle swarm optimization algorithm for global search, and the unorganized emission leakage source of the marked enterprise is located.Setting enterprise source intensity, the pollution concentration contour map is generated by forward diffusion simulation, and the actual air quality contribution value of enterprise emission to surrounding sensitive receptors is quantified.The present application can accurately trace, and provide scientific regional environmental responsibility report.
Need to check novelty before this filing date? Find Prior Art