电机与控制学报
電機與控製學報
전궤여공제학보
ECTRIC MACHINES AND CONTROL
2015年
4期
94-99
,共6页
变分贝叶斯%容积卡尔曼滤波%自适应%非线性系统
變分貝葉斯%容積卡爾曼濾波%自適應%非線性繫統
변분패협사%용적잡이만려파%자괄응%비선성계통
variational Bayes%cubature Kalman filtering%adaptive%non-linear system
针对应用于非线性系统模型的容积卡尔曼滤波工作性能会受观测噪声参数变化的影响而降低的问题，提出一种自适应的变分贝叶斯容积卡尔曼滤波算法。在每一次更新步骤中，将系统状态与变化的观测噪声统计信息一起作为随机变量，并用变分贝叶斯方法进行估计，在迭代逼近得到噪声方差后，再利用容积卡尔曼滤波对系统状态进行更新。仿真实验证明变分贝叶斯容积卡尔曼滤波算法在非线性系统的滤波问题中能够较好跟踪变化的观测噪声方差，相比容积卡尔曼滤波拥有较好的估计性能。
針對應用于非線性繫統模型的容積卡爾曼濾波工作性能會受觀測譟聲參數變化的影響而降低的問題，提齣一種自適應的變分貝葉斯容積卡爾曼濾波算法。在每一次更新步驟中，將繫統狀態與變化的觀測譟聲統計信息一起作為隨機變量，併用變分貝葉斯方法進行估計，在迭代逼近得到譟聲方差後，再利用容積卡爾曼濾波對繫統狀態進行更新。倣真實驗證明變分貝葉斯容積卡爾曼濾波算法在非線性繫統的濾波問題中能夠較好跟蹤變化的觀測譟聲方差，相比容積卡爾曼濾波擁有較好的估計性能。
침대응용우비선성계통모형적용적잡이만려파공작성능회수관측조성삼수변화적영향이강저적문제，제출일충자괄응적변분패협사용적잡이만려파산법。재매일차경신보취중，장계통상태여변화적관측조성통계신식일기작위수궤변량，병용변분패협사방법진행고계，재질대핍근득도조성방차후，재이용용적잡이만려파대계통상태진행경신。방진실험증명변분패협사용적잡이만려파산법재비선성계통적려파문제중능구교호근종변화적관측조성방차，상비용적잡이만려파옹유교호적고계성능。
Focusing on the performance of Cubature Kalman filtering may be degraded due to the fact that in practical situations the statistics of measurement noise might change. An adaptive variational Bayesian cubature Kalman filtering algorithm was proposed which can be used in non-linear system models. In each update step of proposed method, both system state and time-variant measurement noise were recog-nized as random variables to estimate. Measurements noise variances were approximated by variational Bayes, thereafter, system states were updated by cubature Kalman filtering. Simulation results demon-strate the proposed filter can well track measurement noise for a non-linear system and outperforms cuba-ture Kalman filter.
