MENGLAN: Multiscale Enhanced Nonparametric Gas Analyzer with Lightweight Architecture and Networks

TL;DR

MENGLAN is a lightweight, real-time gas analyzer that uses multiscale and attention mechanisms to accurately detect ethylene in mixed gases, improving performance and deployability over traditional methods.

Contribution

The paper introduces MENGLAN, a novel multiscale nonparametric gas analyzer with a hybrid attention mechanism and feature reactivation, enhancing accuracy and efficiency.

Findings

Achieves high-precision ethylene detection in real-time

Reduces computational complexity compared to existing methods

Demonstrates superior performance and deployability

Abstract

Accurate detection of ethylene concentrations in mixed gases is crucial in chemical production for safety and health purposes. Traditional methods are hindered by high cost and complexity, limiting their practical application. This study proposes MENGLAN, a Multiscale Enhanced Nonparametric Gas Analyzer that integrates a dual-stream structure, a Hybrid Multi-Head Attention mechanism, and a Feature Reactivation Module to enable real-time, lightweight, and high-precision ethylene concentration prediction. Results show that MENGLAN achieves superior performance, reduced computational demand, and enhanced deployability compared to existing methods.