小波包和Apriori算法燃气调压器故障诊断与预警

Fault Diagnosis and Early Warning of Gas Regulator Based on Wavelet Packet and Apriori Algorithm

提出一种基于小波包和Apriori算法的燃气调压器故障诊断与预警模型。首先通过小波包算法分析调压器出口压力、负载压力,得出各故障类型的频带能量值,然后采用皮尔逊相关系数法计算出口压力与工况流速、工况流速与阀位相对开度、出口压力与负载压力的相关系数,经事务型转换后应用Apriori算法求得多变量相关系数与故障类型之间的强关联规则。采用小波包算法对调压器出口压力及负载压力进行分解得到的频带能量值可作为基础判据。Apriori算法得出的多变量相关系数与故障类型之间的强关联规则可以作为关键判据,提升高-中压燃气调压器故障诊断与预警模型的精准度。应用小波包算法和Apriori算法相结合的方法建立的高-中压燃气调压器故障诊断与预警模型经实例验证可行。

A fault diagnosis and early warning method of gas regulators based on wavelet packet algorithm and Apriori algorithm is proposed.Firstly,the wavelet packet algorithm is used to analyze the outlet pressure and load pressure of the regulator,and the frequency band energy energy value of each fault type are obtained.Then,the Pearson correlation coefficient method is used to calculate the correlation coefficients between outlet pressure and operating velocity,operating velocity and valve position opening,outlet pressure and load pressure.After transaction type conversion,the Apriori algorithm is used to obtain the strong association rules between multivariable correlation coefficient and fault types.The frequency band energy value obtained by using the wavelet packet algorithm to decompose the outlet pressure and load pressure of the regulator can be used as the basic criterion.The strong association rules between multivariate correlation coefficient and fault type obtained by Apriori algorithm can be used as the key criterion to improve the accuracy of fault diagnosis and early warning model of high and medium pressure gas regulator.The fault diagnosis and early warning model of high and medium pressure gas regulator established by combining wavelet packet algorithm and Apriori algorithm has been verified by practical examples.