Abstract：The larger aircraft cabin air pressure range had great influence on the gas sensor, resulting in inaccurate air quality detect results, RBF neural network was proposed to compensate air pressure. Firstly, the experimental system was designed. Then positive and negative pressure experiments were carried out on four typical gas sensors for cabin air quality detection including HCHO, CO, CO2 and NO2. The test data were collected and the characteristic curves of each gas were drawn. Finally, a three-layer RBF neural network model with 12 pressure points and measured values as inputs and expected values as outputs was established, and the error correction compensation was made to the experimental data. The results showed that the RBF neural network compensation algorithm can reduce the maximum relative error of HCHO, CO, CO2 and NO2 gas sensors from 32.85%, 28.42%, 52.87%, 87.18% to 2.001%, 3.668%, 2.392%, 12.68% respectively, achieve a better compensation effect.
［1］American Society for Testing Material. ASTM D6399-2010 standard guide for selecting instruments and methods for measuring air quality in aircraft cabins ［S］. 2010.
［2］AECMA. BS EN 4618:2009 Aerospace series. Aircraft internal air quality standards, criteria and determination methods［S］. 2009.
［3］国家技术监督局. GB 18883-2002 室内空气质量标准［S］. 北京: 中国标准出版社, 2002.
［4］中华人民共和国卫生部. GB/T 18204-2000公共场所卫生标准检验方法［S］. 北京: 中国标准出版社, 2000.
［5］Haghighat F, Allard F, Megri A C, et al. Measurement of Thermal Comfort and Indoor Air Quality Aboard 43 Flights on Commercial Airlines ［J］. Indoor and Built Environment, 1999, 8(1): 58-66.
［6］Giaconia C, Orioli A, Gangi A D. Air quality and relative humidity in commercial aircrafts: An experimental investigation on short-haul domestic flights［J］. Building and Environment, 2013, 67(3): 69-81.
［7］代炳荣. 飞机座舱动态空气净化系统研究［D］. 南京:南京航空航天大学, 2014.
［8］Li M X , Zhao B, Tu J Y, et al. Study on the carbon dioxide lockup phenomenon in aircraft cabin by computational fluid dynamics ［J］. Building Simulation, 2015, 8(4): 431-441.
［9］Zhao Y J, Dai B R, Yu Q, et al. Numerical simulation study on air quality in aircraft cabins ［J］. Journal of Environmental Sciences, 2017, 56(6): 52-61.
［10］张小俊, 张明路, 李小慧. 基于RBF神经网络的电化 学CO气体传感器的温度补偿［J］. 传感技术学报, 2009, 22(1): 11-14.
Zhang X J, Zhang M L, Li X H. Temperature compensation of electrochemical CO gas sensor based on RBF neural network ［J］. Journal of sensor technology, 2009, 22(1): 11-14.
［11］张淑清,任爽,陈荣飞,等. 基于大数据简约及PCA改进RBF网络的短期电力负荷预测［J］. 计量学报, 2018, 39(3): 392-396.
Zhanh S Q, Ren S, Chen R F, et al. Short-term Power Load Forecast Based on Big Data Reduction and PCA-improved RBF Network［J］. Acta Metrologica Sinica, 2018, 39(3): 392-396.
［11］陆艺, 张培培, 王学影, 等. 基于PSO-BP神经网络的关节臂式坐标测量机长度误差补偿［J］. 计量学报, 2017, 38(3): 271-275.
Lu Y, Zhang P P, Wang X Y, et al. Error compensation of joint arm coordinate measurement based on PSO-BP neural network ［J］. Acta Metrologica Sinica, 2017, 38(3): 271-275.
［12］王有贵, 吴双双, 陈红江. 称重传感器蠕变误差的神经网络补偿方法［J］. 计量学报, 2018, 39(4): 510-514.
Wang Y G, Wu S S, Chen H J. Neural network compensation method for creep error of weight sensor ［J］. Acta Metrologica Sinica, 2018, 39(4): 510-514.
［13］Poggio T, Girosi F. Networks for approximation and learning ［J］. Proceedings of the IEEE, 1990, 78(9): 1481-1497
［14］何平, 潘国峰, 赵红东, 等. 基于RBF网络的智能气敏传感器温度补偿［J］. 仪表技术与传感器, 2008,(7): 6-8+42.
He P, Pan G F, Zhao H D, et al. Temperature compensation of intelligent gas sensor based on RBF network ［J］. Instrument technology and sensor, 2008,(7): 6-8+42.
2020, Vol. 41 
