On guaranteed parameter estimation of a multiparameter linear regression process 1;2 we u uhler yheslv eF siliev snstitute of wthemtis heprtment of epplied wthemtis rumoldt niversity ferlin nd gyernetis nter den vinden TD hEIHHWW omsk tte niversity ferlinD qermny venin QTD TQRHSH omskD ussi Abstract his pper presents sequentil estimtion proedure for the unknown prmeters of ontinuousEtime stohsti liner regression proessF es exmples the sequenE til estimtion prolem of two dynmi prmeters in

more »

... liner systems with memory nd in utoregressive proesses is solvedF he estimtion proedure is sed on the lest squres method with weights nd yields estimtors with gurnteed E ury in the sense of the L q norm for xed q ! PF he proposed proedure works in the mentioned exmples for ll possile vlues of unknown dynmi prmeters on the plne R P for the utoregressive proesses nd on the plne R P with the exeption of some lines for the liner stohsti dely equtionsF he symptoti ehvior of the durtion of oservtions is determinedF he generl estimtion proedure is designed for twoE or moreEprmetri modelsF st is shownD tht the proposed proedure n e pplied to the sequentil prmeter estimtion prolem of ne stohsti dely dierentil equtions nd utoregresE sive proesses of n ritrry orderF