Run estimate(Mdl,y) with constraints on coefficients of ssm(A,B,C,D) model

1 次查看（过去 30 天）

显示更早的评论

Oli 2016-3-24

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/275264-run-estimate-mdl-y-with-constraints-on-coefficients-of-ssm-a-b-c-d-model

编辑： Oli 2016-3-24

Hi all,

I would like to run following function to estimate the coefficients for an ssm model:

EstMdl = estimate(Mdl,y,'name','value')

and an ssm model of following form:

Mdl = ssm(A,B,C,D)

These vectors contains 3 undefined coefficients which are named 'NaN'. A = [NaN NaN; 0 0]; D = NaN;

My questions: 1) How can I introduce the constraint that coefficient 1 and 2 (those of vector A) have to sum up to 1 (NaN(1)+NaN(2)=1). I assume something like 'beq' could work as a name/value pair, but I could not find a decent formula? 2) Can I give the NaN coefficients different names (e.g. 'w','v') in order to impose constraints such as w+v=1

Many thanks for your advice.

Oli