| In recent years,with the increasing volume of natural gas trade in China,it is more and more important to ensure the accuracy and reliability of natural gas measurement.As a secondary standard in the process of flowmeter calibration,the accuracy of the critical flow venturi nozzle method gas flow standard device is an important guarantee of the accuracy of the working level standard device.At present,the critical flow venturi nozzle method gas flow standard device quality(uncertainty)is mainly used during the verification,repeatability and stability testing off-line methods to ensure,but these methods have a lag.In order to solve this problem,the concept of equipment integrity management is introduced,and the quality control study of this gas flow standard device is carried out for the critical flow venturi nozzle method gas flow standard device of a national high flow oil and gas metering station.Firstly,we identify the factors affecting the quality of the critical flow venturi nozzle gas flow standard device.Based on the calibration principle of the standard device,20 influencing factors were identified,among which the key influencing factors were the accuracy of temperature transmitter,pressure transmitter and chromatographic analyzer;on this basis,multivariate differentiation was used to calculate the deviation of the value transfer when the accuracy of the measuring instrument deviated from the size,and the results showed that the deviation of pressure and gas component deviated more than the temperature deviation to bring about the relative deviation of flow.Next,a risk evaluation of the critical flow venturi nozzle method gas flow standard device was performed.The calculation method of the risk value was determined by combining the risk-based test;the error analysis of random forest,LSTM,fully connected layer neural network,SVM and ARIMA algorithms was performed,and the random forest model was selected as the prediction algorithm for the probability calculation of trend extrapolation method,and the calculation method of the risk probability of quantity transfer was derived by combining the trend extrapolation method and the historical data method;the mean-standard deviation-based Control method based on the mean-standard deviation method to study the calculation of the risk consequences of the transfer of secondary standard devices;the combination of the probability of the transfer of risk and the consequences of risk,the unified risk value of different measuring instruments scoring,a pressure transmitter as an example of risk evaluation,the evaluation results show that the pressure transmitter risk value of 0.1,the risk is low,the disposal measures for normal maintenance use.Finally,the quality control measures are optimized for the lagging problem of the existing control measures of substandard devices.A multi-pipe combination prediction algorithm was studied based on neural network,so as to monitor the device online,but there are limitations in practical application due to the small amount of data;a method based on multivariate differential deviation calculation was proposed independently,and the calculation results showed that the method can accurately calculate the deviation of the quantity-value transfer,which has a certain degree of generalizability;combined with the evaluation method of the quantity-value transfer risk,the quality control measures of the substandard device were summarized,and the substandard device did not suffer from lag after using the measures.The uncertainty of the secondary standard device can be guaranteed within 2‰ after using this measure,which can achieve the purpose of quality control. |
PDF Full Text Request