期刊文献+
共找到3篇文章
< 1 >
每页显示 20 50 100
基于加权时变泊松模型的出租车载客点推荐模型 被引量:6
1
作者 商建东 李盼乐 +1 位作者 刘润杰 李润川 《计算机应用》 CSCD 北大核心 2018年第4期923-927,934,共6页
针对出租车空载率高、司机寻客难的问题,提出泊松-卡尔曼组合预测模型(PKCPM)。首先,采用加权非齐次泊松模型,针对出租车历史数据进行建模,得到目标时刻的估计值;其次,基于当天的实时数据,将临近时刻乘客需求的平均值作为目标时刻预测值... 针对出租车空载率高、司机寻客难的问题,提出泊松-卡尔曼组合预测模型(PKCPM)。首先,采用加权非齐次泊松模型,针对出租车历史数据进行建模,得到目标时刻的估计值;其次,基于当天的实时数据,将临近时刻乘客需求的平均值作为目标时刻预测值;最后,将预测值和估计值作为卡尔曼滤波模型的输入参数,实现对目标时刻出租车乘客需求的预测,同时引入误差反向传播机制,减小下一次预测误差。基于郑州市出租车轨迹数据集,对组合模型与非齐次泊松模型(NHPM)、加权非齐次泊松模型(WNHPM)、支持向量机(SVM)等三种模型进行对比,实验结果显示PKCPM的误差比WNHPM、SVM分别降低了8.85个百分点、14.9个百分点。该模型能对不同时段内、不同空间网格的乘客需求进行预测,为出租车寻找乘客提供可靠的依据。 展开更多
关键词 空载率 卡尔曼滤波预测模型 加权时变泊松模型 临近时刻乘客需求
下载PDF
MPF-CKF在SINS大方位失准角初始对准中的应用
2
作者 贾鹤鸣 宋文龙 +1 位作者 牟宏伟 车延庭 《北京工业大学学报》 CAS CSCD 北大核心 2013年第10期1468-1473,共6页
针对捷联惯导系统在大方位失准角情况下的初始对准问题,提出了一种基于MPF-CKF的非线性滤波方法.MPF-CKF将部分惯性器件误差作为模型误差,降低了系统的维数,不仅提高了初始对准的精度,而且克服了将模型误差假设为高斯白噪声的局限性.通... 针对捷联惯导系统在大方位失准角情况下的初始对准问题,提出了一种基于MPF-CKF的非线性滤波方法.MPF-CKF将部分惯性器件误差作为模型误差,降低了系统的维数,不仅提高了初始对准的精度,而且克服了将模型误差假设为高斯白噪声的局限性.通过滤波仿真比较,进一步表明了MPF-CKF能提高SINS在大方位失准角初始对准中的估计精度和收敛速度. 展开更多
关键词 捷联惯导 大方位失准角 初始对准 模型预测滤波-容积卡尔曼滤波(MPF—CKF)
下载PDF
Empirical modeling of ionospheric F2 layer critical frequency over Wakkanai under geomagnetic quiet and disturbed conditions 被引量:4
3
作者 LIU Jing LIU LiBo +2 位作者 ZHAO BiQiang WAN WeiXing CHEN YiDing 《Science China(Technological Sciences)》 SCIE EI CAS 2012年第5期1169-1177,共9页
The hourly values of the ionospheric F2 layer critical frequency, foF2, recorded at Wakkanai ionosonde station (45.4°N, 141.7°E) have been collected to construct a middle-latitude single-station model for ... The hourly values of the ionospheric F2 layer critical frequency, foF2, recorded at Wakkanai ionosonde station (45.4°N, 141.7°E) have been collected to construct a middle-latitude single-station model for forecasting foF2 under geomagnetic quiet and disturbed conditions. The module for the geomagnetic quiet conditions incorporates local time, seasonal, and solar vari- ability of climatological foF2 and its upper and lower quartiles. It is the first attempt to predict the upper and lower quartiles of foF2 to account for the notable day-to-day variability in ionospheric foF2. The validation statistically verifies that the model captures the climatological variations of foF2 with higher accuracy than IRI does. The storm-time module is built to capture the geomagnetic storm induced relative deviations of foF2 from the quiet time references. In the geomagnetically disturbed module, the storm-induced deviations are described by diumal and semidiumal waves, which are modulated by a modified magnetic activity index, the Kf index, reflecting the delayed responses of foF2 to geomagnetic activity forcing. The coeffi- cients of the model in each month are determined by fitting the model formula to the observation in a least-squares way. We provide two options for the geomagnetic disturbed module, including or not including Kalman filter algorithm. The Kalman filter algorithm is introduced to optimize these coefficients in real time. Our results demonstrate that the introduction of the Kalman filter algorithm in the storm time module is promising for improving the accuracy of predication. In addition, comparisons indicate that the IRI model prediction of the F2 layer can be improved to provide better performances over this region. 展开更多
关键词 Empirical modeling Kalman f'dter ionospheric storm
原文传递
上一页 1 下一页 到第
使用帮助 返回顶部