Research On The Information Fusion For Air Pressure Detection Of Electric Gas Pressure Regulator

In recent years, the development and utilization of natural gas plays an importantrole in China’s economic development and environmental protection. As one of themajor equipment in city’s gas transmission and distribution system, a gas pressureregulator is the pivotal link to achieve high quality and efficient use of natural gasresources. Traditional gas regulator generally uses the mechanical controlled principle.The regulating accuracy is not high, so that gas cannot be sufficiently burned. It cangive rise to energy waste and environmental pollution, and even induce gas securityissues. In order to fundamentally solve the existing problems of the mechanical gasregulator, electric pressure regulating is the main trend for future regulator’sdevelopment. In the electric regulating model, accurate pressure detection of the gasregulator’s inlet and outlet plays the key role in improving the overall performance forproviding the reliable original data for finally realizing adaptive regulating.On the basis of in depth studying the existing pressure detection method, theworking principle of a piezoresistive pressure sensor and other related technologies, thedetection scheme of multi-source information fusion based on an intelligent air pressuresensor array is proposed, combining with the actual application requirements of thepressure detection for the electric gas pressure regulator’s inlet and outlet. We concretelyanalyze the construction strategy of the sensor array, and focus on the informationfusion algorithm of the air pressure detection system, to improve accuracy andreliability of the pressure detection.The main work of the paper is as follows:①Structuring the intelligent air pressure sensor array for the pressure detection ofthe electric gas pressure regulator’s inlet and outlet. On the basis of in depth studyingthe basic structure and working principle of a piezoresistive pressure sensor, we analyzethe inherent static characteristics of a pressure sensor, to construct the intelligent airpressure sensor with temperature compensation. Considering practical needs withvolume of a single intelligent air pressure sensor and cross-sectional area of theregulator’s inlet and outlet, we construct4×4intelligent air pressure sensor array byusing the intelligent air pressure sensor as basic elements.②Researching on the information fusion algorithm of the pressure detection forthe electric gas pressure regulator’s inlet and outlet. According to the characteristics of the intelligent air pressure sensor array, the information fusion algorithm for thepressure detection of the electric gas pressure regulator’s inlet and outlet is proposed,further researching on the algorithm of RBF neural network and the adaptive weightedalgorithm. We use the modified dynamic nearest neighbor-clustering algorithm to trainthe RBF neural network to complete the heterogeneous sensor information fusion andthe adaptive weighted algorithm to complete the homogeneous sensor informationfusion.③Completing simulation and analysis for the information fusion algorithm of theintelligent air pressure sensor array. Three kinds of the heterogeneous sensorinformation fusion algorithms are simulated and compared transversely. On the basis,the homogeneous sensor information fusion algorithm and they are analyzed andcompared longitudinally, to verify the performance of the information fusion algorithmin the paper.Theoretical analysis and simulation results show that the proposed detectionscheme of multi-source information fusion based on the intelligent air pressure sensorarray can improve accuracy and reliability in detecting gas regulator’s inlet and outletpressure. It provides strong technical support for electric gas regulating.