称重雨量数据处理卡尔曼滤波应用

Application of Kalman Filter in Processing Weighing Rain Data

降雨量的测量,目前业务上应用较为广泛的是翻斗式雨量计,它只能测降雨,对于冰雹、降雪等固态降水的测量采用人工观测为主,称重式雨量计与翻斗式雨量计相比,其优势在于能实现所有类型降水的全天候自动化观测。本研究随机选择了一天无降水数据确定了滤波参数Q和R(过程噪声方差和观测噪声方差),根据确定的滤波参数,随机选择了无降水(2016年4月3日)和有降水(2015年7月21日、2015年8月7日)日采用卡尔曼滤波,并结合翻斗雨量传感器数据进行验证,结果表明,本研究确定的滤波参数采用卡尔曼滤波后能够有效去除称重雨量中的噪声,使滤波后的曲线变得平稳光滑,减小了数据的抖动频率和误差。

The tipping-bucket rain-gauge is widely used in meteorological services at present to measure rainfall,which can only measure rainfall.For measuring hail,snow and other solid precipitation,the manual method is used mainly.Comparing with the tipping-bucket rain gauge,the advantage of the weighing rain gauge is to achieve the automatic observation of various types of precipitation.The study randomly selected a day without precipitation data to determine the filter function Qand Rvalues,based on the determined filter parameters,randomly selected a non-precipitation day（3 April 2016） and two precipitation days（21 July 2015,7 August 2015） using the Kalman filter,combined with tipping-bucket sensor data validation,the result shows that the filter parameters determined by using the Kalman filter can effectively remove the weighing rain noise,so that the filtered curve becomes smooth and steady,and reduce the data frequency jitters and errors.