Research On Early Warning Expert System For Gas Based On SVM

Gas accidents are currently the biggest threat and most notably problem to coal mine safety in production. Accurately early warning of gas is of great theoretical and practical significance. Meanwhile, because of the factors that affect gas outburst with uncertainty and ambiguity,when finding the potential gas outburst or having happened gas accidents, taking timely and effective measures is a realistic problem.Support Vector Machine is a new method of machine learning. It bases on the statistical learning theory, and can settle"small"example problem well. Because of its excellent learning ability, SVM has been applied to many fields. Case-based Reasoning, because of it overcoming the traditional rule-based systems many shortcomings, such as knowledge acquisition bottleneck problem, to the processed problem had no memory, the overall performance of more vulnerable and so on, is more and more paid attention and widely used in various fields.In this paper, Support Vector Machine theory is applied for gas outburst early warning and case-based reasoning theory is applied for gas accidents case-based reasoning. First, we preprocess the data obtained from sentinel surveillance, extract characteristic parameters and normalize them, use Support Vector Machine for training and classifying the outburst area, by adjusting the coefficient of real-time early warning alarm; At the same time, we preserve cases of prominent warning and cases happened explosion accident , and set up accident case database. And then we match searching the current accident case in the accident case database.Then we put the similar case as the starting point for dealing with the current accident to develop treatment plans in order to reduce accident loss as much as possible.Because of the contents of this paper is based on the actual project, this paper gives a detailed theoretical basis and concrete solutions on data preprocessing, feature extraction, support vector classification, the adjustment factor to early warning and the implementation of gas accidents case-based reasoning. The arrangement of the thesis overall chapters is the model of summary presentation, theoretical basis, practical application. This paper is based on the practical application of coal mine design, and the algorithms and design ideas of this paper are general.Therefore, they can also be used for other systems of treatment and application of fault diagnosis.