哈尔滨工程大学学报
哈爾濱工程大學學報
합이빈공정대학학보
JOURNAL OF HARBIN ENGINEERING UNIVERSITY
2014年
5期
630-636
,共7页
机动目标跟踪%变结构多模型方法%多因素模糊综合判决%非线性复杂系统%状态估计%MSA 策略
機動目標跟蹤%變結構多模型方法%多因素模糊綜閤判決%非線性複雜繫統%狀態估計%MSA 策略
궤동목표근종%변결구다모형방법%다인소모호종합판결%비선성복잡계통%상태고계%MSA 책략
maneuvering target tracking%variable structure multi-model method%multi-factor fuzzy synthetic evalua-tion%nonlinear hybrid system%state estimation%MSA strategy
为提高非线性复杂系统状态估计的效率与精度，提出一种基于多因素模糊综合评判的变结构多模型方法（ MFIE MM）。MFIE MM 首先确定模型全集并提取各个模型的公共因素，进而选择模糊综合鉴别函数构建模糊评价集合；其次单因素模糊评判矩阵和多因素模糊评判准则得到各个模型的相似度；最后选出当前时刻最佳模型并以此模型为区域中心实时生成参与状态估计的模型集合。仿真结果显示，由 MFIE MM 处理得到的位置变量估计误差协方差从2.15 m降到2.05 m，单拍处理时间从0.0027 s 降到0.0018 s。因此，MFIE MM 在显著提高算法精度的同时有效降低了算法运行时间和模型平均误差。
為提高非線性複雜繫統狀態估計的效率與精度，提齣一種基于多因素模糊綜閤評判的變結構多模型方法（ MFIE MM）。MFIE MM 首先確定模型全集併提取各箇模型的公共因素，進而選擇模糊綜閤鑒彆函數構建模糊評價集閤；其次單因素模糊評判矩陣和多因素模糊評判準則得到各箇模型的相似度；最後選齣噹前時刻最佳模型併以此模型為區域中心實時生成參與狀態估計的模型集閤。倣真結果顯示，由 MFIE MM 處理得到的位置變量估計誤差協方差從2.15 m降到2.05 m，單拍處理時間從0.0027 s 降到0.0018 s。因此，MFIE MM 在顯著提高算法精度的同時有效降低瞭算法運行時間和模型平均誤差。
위제고비선성복잡계통상태고계적효솔여정도，제출일충기우다인소모호종합평판적변결구다모형방법（ MFIE MM）。MFIE MM 수선학정모형전집병제취각개모형적공공인소，진이선택모호종합감별함수구건모호평개집합；기차단인소모호평판구진화다인소모호평판준칙득도각개모형적상사도；최후선출당전시각최가모형병이차모형위구역중심실시생성삼여상태고계적모형집합。방진결과현시，유 MFIE MM 처리득도적위치변량고계오차협방차종2.15 m강도2.05 m，단박처리시간종0.0027 s 강도0.0018 s。인차，MFIE MM 재현저제고산법정도적동시유효강저료산법운행시간화모형평균오차。
A variable structure multi-model method based on the multi-factor fuzzy integrated evaluation(MFIE MM)is presented to improve the estimation precision and efficiency for nonlinear hybrid systems. MFIE MM ini-tially ascertains the total model set and extracts the common factors of all the models,then chooses the fuzzy syn-thetic discriminant function and constructs the fuzzy evaluation set. After this,the similarity of every model through calculating the single-factor fuzzy judge matrix and the multi-factor fuzzy judge criterion is obtained. Finally it se-lects the best model at the current time and centers on this model as the regional center to produce in time the mod-el set which joins in the state estimation. The simulation results show that the estimation error variances belonging to the position of MFIE MM are improved from 2.15 m to 2.05 m and the time of one cycle is decreased from 0.002 7 s to 0.001 8 s. So a conclusion is derived that MFIE MM evidently improves the precision,shortens the running time of the algorithm,and effectively reduces the average error rate of the model.
