山东农业大学学报(自然科学版)
山東農業大學學報(自然科學版)
산동농업대학학보(자연과학판)
JOURNAL OF SHANDONG AGRICULTURAL UNIVERSITY(NATURAL SCIENCE)
2014年
2期
204-209
,共6页
汽车行驶姿态%预测%AR模型%MLR模型
汽車行駛姿態%預測%AR模型%MLR模型
기차행사자태%예측%AR모형%MLR모형
Vehicle driving posture%forecast%auto-regressive model%multi-level recursive model
运用自回归(AR)建模预测方法及多层递阶(MLR)建模预测方法对汽车行驶姿态进行预测。在详细分析了两种的建模预测方法的基础上,以汽车的侧倾角速度预测为例讨论了两种建模预测方法在汽车行驶姿态预测中的应用。结果显示AR建模预测方法有一定的预测效果,1 s预测平均相对误差为10.56%,MLR建模预测方法由于充分考虑到了汽车运动的时变性和非线性特性,其预测效果更佳,1 s及2 s的预测精度分别比AR建模预测法提高6.44%和7.09%。
運用自迴歸(AR)建模預測方法及多層遞階(MLR)建模預測方法對汽車行駛姿態進行預測。在詳細分析瞭兩種的建模預測方法的基礎上,以汽車的側傾角速度預測為例討論瞭兩種建模預測方法在汽車行駛姿態預測中的應用。結果顯示AR建模預測方法有一定的預測效果,1 s預測平均相對誤差為10.56%,MLR建模預測方法由于充分攷慮到瞭汽車運動的時變性和非線性特性,其預測效果更佳,1 s及2 s的預測精度分彆比AR建模預測法提高6.44%和7.09%。
운용자회귀(AR)건모예측방법급다층체계(MLR)건모예측방법대기차행사자태진행예측。재상세분석료량충적건모예측방법적기출상,이기차적측경각속도예측위례토론료량충건모예측방법재기차행사자태예측중적응용。결과현시AR건모예측방법유일정적예측효과,1 s예측평균상대오차위10.56%,MLR건모예측방법유우충분고필도료기차운동적시변성화비선성특성,기예측효과경가,1 s급2 s적예측정도분별비AR건모예측법제고6.44%화7.09%。
Auto-regressive model and Multi-level Recursive model were adopted to forecast vehicle motion posture in this paper. The modeling processes of them were represented in detail. Then their application were discussed couple with example of roll velocity forecast. The result showed that AR forecasting method has some predictive effect, 1 second predicted average relative error was 10.56%, and Multi-level Recursive forecasting reached higher accuracy for its consideration of non-linear and time-variant characteristic. 1 second and 2 seconds prediction accuracy compared with the AR model forecasting method improved 6.44%and 7.09%.