欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
针对现有单一导航卫星钟差预报模型存在预报精度不高的问题,提出了改进粒子群算法优化的组合预报模型。该模型利用差分自回归移动平均模型（ARIMA）和最小二乘向量机（LSSVM）模型的特点,首先建立ARIMA模型预报钟差数据的线性部分,并得到预报残差;然后,根据残差建立LSSVM模型预报非线性部分,最后的预报结果即两个预报结果之和。同时引入随优化代数变化的惯性权值和加速度因子,来提高粒子群（PSO）算法寻优能力,并用其优化组合预报模型中LSSVM部分的惩罚因子和核函数参数选取过程,以提高模型的预报精度。实例与结果分析表明,组合模型较单一模型在预报精度上有30%~50%的提高,为导航卫星高精度短期钟差预报提供了一种新思路。
英文摘要：
Aiming at the single model＇s poor performance on navigation satellite clock error＇s predication,a combined method optimized by improved particle swarm was proposed.This model used the characters of ARIMA model and LSSVM model.Firstly,the ARIMA model was established to predict the linear relation of satellite clock error data,and the LSSVM model using the residual sequence was built to compensate the nonlinear law of the clock error data.The result of this combined model was the sum of both the predictions of ARIMA model and LSSVM model.Besides,the improved particle swarm optimization which had the inertia weight and acceleration value changing with the generation had better performance to find the better parameters of model.And it was used to optimize the parameters of LSSVM which belonged to the combined model.The results of the simulation showed that the accuracy of new combined model was superior to the single model with 30%~50% promotion.And the work provides a new way for short term prediction of navigation satellite clock error.
同期刊论文项目
同项目期刊论文
An Adaptive ISAR-Imaging-Considered Task Scheduling Algorithm for Multi-Function Phased Array Radars
期刊信息
位置：