摘要
为提高自动气象站温度传感器观测数据精度,对不同型号的自然通风防辐射罩所产生的辐射误差进行分析,提出一种基于BP神经网络算法对不同表面反射率的防辐射罩进行误差修正。将BP神经网络应用于温度传感器防辐射罩辐射误差的预测:将太阳辐射强度、风速、表面反射率作为BP神经网络的输入,利用计算流体动力学分析防辐射罩在不同大气环境下的辐射误差作为BP神经网络的训练输出。分析训练输出与样本输出,两者的绝对误差仅在[-0. 001,0. 002],可见BP神经网络的预测精度在理想值内。最后将BP神经网络得到的辐射误差修正方程用Java进行封装,并开发Web平台实现算法应用。
In order to improve the accuracy of temperature sensor observation data of automatic weather station,the radiation errors caused by different types of natural ventilation radiation shield were analyzed. A BP neural network algorithm was proposed to correct the errors of radiation shield with different surface reflectivity. The BP neural network is applied to the prediction of radiation error of the radiation shield of the temperature sensor: the solar radiation intensity,wind speed and surface reflectivity are used as the inputs,and the calculation of the radiation error of the radiation hood in different atmospheric environment is used as the output in the training of the BP neural network. The absolute error of training output and sample output is in the interval [-0. 001,0. 002],which shows that the prediction accuracy of BP neural network is within the ideal value. Finally,the radiative error correction equation obtained by BP neural network is encapsulated by Java,and the Web platform is developed to implement the algorithm.
作者
浦玮
刘清惓
史雪雪
王定奥
PU Wei;LIU Qingquan;SHI Xuexue;WANG Dingao(College of Electronics and Information Engineering, Nanjing University of Information Science and Technology, Nanjing 210044, China;Jiangsu Key Laboratory of Meteorological Observation and Information Processing, Nanjing University of Information Science and Technology, Nanjing 210044, China;Jiangsu Collaborative Innovation Center on Atmospheric Environment and Equipment Technology, Nanjing University of Information Science and Technology, Nanjing 210044, China)
出处
《实验室研究与探索》
CAS
北大核心
2019年第4期13-16,37,共5页
Research and Exploration In Laboratory
基金
国家自然科学基金项目(4127505042)
国家公益性行业(气象)科研专项项目(GYHY200906037
GYHY201306079)
江苏高校优势学科Ⅱ期建设工程资助项目(PAPD-Ⅱ)
