A Novel Optimization Technique to Improve Gas Recognition by Electronic Noses Based on the Enhanced Krill Herd...

Area of Science:

• Artificial Intelligence
• Environmental Monitoring
• Chemical Sensing

Background:

• Electronic noses (E-noses) are intelligent systems for gas detection.
• Accurate identification of indoor pollutants like benzene, toluene, formaldehyde, and carbon monoxide is crucial.
• Support Vector Machine (SVM) is a common pattern recognition algorithm for E-nose data, but its performance depends on parameter optimization.

Purpose of the Study:

• To enhance the gas recognition rate of an E-nose system.
• To optimize the parameters of a Support Vector Machine (SVM) algorithm for improved indoor pollutant detection.
• To introduce an effective Enhanced Krill Herd (EKH) algorithm for SVM parameter optimization.

Main Methods:

• Developed an Enhanced Krill Herd (EKH) algorithm with a novel decision weighting factor and updated crossover operator.
• Utilized EKH to optimize two key parameters of the Support Vector Machine (SVM) for gas recognition.
• Compared the performance of EKH against other optimization algorithms (KH, CKH, QPSO, PSO, GA) using E-nose data.

Main Results:

• The proposed EKH algorithm demonstrated superior performance in optimizing SVM parameters compared to other methods.
• EKH achieved a higher gas recognition rate for indoor pollutants (benzene, toluene, formaldehyde, CO) than conventional algorithms.
• The enhanced krill behavior model in EKH improved global searching and convergence speed.

Conclusions:

• The Enhanced Krill Herd (EKH) algorithm significantly improves the performance and accuracy of electronic nose systems for indoor air quality monitoring.
• EKH provides a robust and effective method for optimizing SVM parameters in gas recognition tasks.
• This study lays the groundwork for further research into improved krill algorithms for diverse E-nose applications.