Gas Concentration Monitoring Prewarning Based on Adaptive Prediction and Feature Extraction

Table 3

Comparison with other prewarning methods.

Relevant research	Prewarning model	Implementation process	Prewarning reliability
[10]	Subjective indicator prewarning	Subjectively set the threshold of indicators derived from regulations	The determination of threshold is subjective, and abnormality is not recognized
[11]	Probability prewarning	Prewarning analysis based on statistical parameters	Uses a single statistical analysis and fails to reliably identify anomalies
[12]	Numerical calculation prewarning	Set the threshold of indicators derived from regulations by numerical calculation	The threshold is fixed, and the reliability is unknown
[13]	Numerical calculation prewarning	Compare the results of the abnormal analysis with the alarm values derived from regulations	Using the alarm value as the threshold, abnormality is not reasonably recognized
[14]	Probability prewarning	Prewarning analysis based on statistical parameters	Only statistical values are linearly superimposed, and the reliability is unknown
	This research method	Dynamically determine the prewarning threshold based on the adaptive prediction and the quantitative analysis of the production process impact	The abnormal conditions are reasonably and reliably identified