Research On The Infrared Methane Detecting System Based On Support Vector Machine Used In Mine

The infrared absorption principle and produce mechanism are illustrated. The exiting problem of infrared absorption band model is also indicated in the paper. In order to compensate the errors from interference factors, one light and two detectors optical system based on reflection and two light and two detectors optical system based on spectrophotometry are designed. The two light and two detectors optical system based on spectrophotometry is effective of capacitive proved by theoretical calculation and the experimental result. Support vector machine based on statistical learning theory is powerful for the problem characterized by small sample, nonlinearity and local minima. For the sake of the limitation of the ability for errors compensation of optical system and the infrared absorption band model, the infrared methane sensor mathematical model based on support vector regression is built to enhance the detection precision of methane. The structure of the model is established after training and learning. And the control capability of model parameter and kernel is also researched thoroughly. The veracity and validity of the infrared methane sensor mathematical model based on support vector machine are confirmed by the experimentation. A new style of infrared methane sensor and detecting system based on embed Linux real time system are designed used in coal mine, and the prototype is completed. The experiment outcomes indicate the satisfied capability of the new instrument, that is anti-jamming, high stability and precision, low power consumed,and so on, which meet the demand of coal mine.