 The Orocos ProjectSmarter control in robotics & automation! |
|
Signal filters defined in continuous domain
Hi, I am looking for a library that allows to use signal filters like low-pass filters, band-pass filters,... defined in the continuous domain. This means that you specify the filter type, the low and/or high bandwidth, the order of the filter, the discretization method (ZOH, matched, bilinear,...) and that's it. This can be very useful while tuning a system. Any suggestions? Thanks, Bert Signal filters defined in continuous domainSubmitted by phamelin on Thu, 2010-11-18 03:40.
2010/11/17 <bert [dot] willaert [..] ...> > On Wed, Nov 17, 2010 at 02:58, > bert [dot] willaert [..] ... > <bert [dot] willaert [..] ...> wrote: > >> Hi, > >> > >> I am looking for a library that allows to use signal filters like > low-pass filters, band-pass filters,... defined in the continuous domain. > This means that you specify the filter type, the low and/or high bandwidth, > the order of the filter, the discretization method (ZOH, matched, > bilinear,...) and that's it. This can be very useful while tuning a system. > Any suggestions? > > A few days ago, we bumped into aquila (http://aquila-dsp.org/), it > says it's a c++ library for digital signal processing so I have no > idea if it has a lot of functionality for the continuous domain. > > And just wondering: could you not use discretization? In practice you > will have a discrete system in the end anyway a suppose. > > Tinne > > In the end you have a discrete system that 's right, but while tuning a > position loop e.g. I normally try a lot of different filter settings. Not > having to calculate the discrete equivalent everytime (with matlab/octave > e.g.) and change the coefficients of the discrete tf could save 'the tuner' > a lot of time. Also when changing the sample time you have to change the > coefficients of the discrete tf... > > In that case, I usually code my algorithm in the discrete-time form, but the specification are input in continuous-time domain. The casting of parameters (specifications) from continuous to discrete-time is done inside the C++ code. > > > -- > Orocos-Users mailing list > Orocos-Users [..] ... > http://lists.mech.kuleuven.be/mailman/listinfo/orocos-users > Signal filters defined in continuous domainSubmitted by b1willaert on Wed, 2010-11-17 02:00.
Hi, I am looking for a library that allows to use signal filters like low-pass filters, band-pass filters,... defined in the continuous domain. This means that you specify the filter type, the low and/or high bandwidth, the order of the filter, the discretization method (ZOH, matched, bilinear,...) and that's it. This can be very useful while tuning a system. Any suggestions? Thanks, Bert |
Signal filters defined in continuous domain
On Wed, Nov 17, 2010 at 02:58, bert [dot] willaert [..] ...
<bert [dot] willaert [..] ...> wrote:
> Hi,
>
> I am looking for a library that allows to use signal filters like low-pass filters, band-pass filters,... defined in the continuous domain. This means that you specify the filter type, the low and/or high bandwidth, the order of the filter, the discretization method (ZOH, matched, bilinear,...) and that's it. This can be very useful while tuning a system. Any suggestions?
A few days ago, we bumped into aquila (http://aquila-dsp.org/), it
says it's a c++ library for digital signal processing so I have no
idea if it has a lot of functionality for the continuous domain.
And just wondering: could you not use discretization? In practice you
will have a discrete system in the end anyway a suppose.
Tinne
>
> Thanks,
>
> Bert
>
>
> --
> Orocos-Users mailing list
> Orocos-Users [..] ...
> http://lists.mech.kuleuven.be/mailman/listinfo/orocos-users
>
Signal filters defined in continuous domain
<bert [dot] willaert [..] ...> wrote:
>> Hi,
>>
>> I am looking for a library that allows to use signal filters like low-pass filters, band-pass filters,... defined in the continuous domain. This means that you specify the filter type, the low and/or high bandwidth, the order of the filter, the discretization method (ZOH, matched, bilinear,...) and that's it. This can be very useful while tuning a system. Any suggestions?
says it's a c++ library for digital signal processing so I have no
idea if it has a lot of functionality for the continuous domain.
And just wondering: could you not use discretization? In practice you
will have a discrete system in the end anyway a suppose.
Tinne
In the end you have a discrete system that 's right, but while tuning a position loop e.g. I normally try a lot of different filter settings. Not having to calculate the discrete equivalent everytime (with matlab/octave e.g.) and change the coefficients of the discrete tf could save 'the tuner' a lot of time. Also when changing the sample time you have to change the coefficients of the discrete tf...
Bert
>
> Thanks,
>
> Bert
>
>
> --
> Orocos-Users mailing list
> Orocos-Users [..] ...
> http://lists.mech.kuleuven.be/mailman/listinfo/orocos-users
>
Re: Signal filters defined in continuous domain
And just wondering: could you not use discretization? In practice you will have a discrete system in the end anyway a suppose.
Tinne
In the end you have a discrete system that 's right, but while tuning a position loop e.g. I normally try a lot of different filter settings. Not having to calculate the discrete equivalent everytime (with matlab/octave e.g.) and change the coefficients of the discrete tf could save 'the tuner' a lot of time. Also when changing the sample time you have to change the coefficients of the discrete tf...
Bert
> > Thanks, >> Bert >> > -- > Orocos-Users mailing list > Orocos-Users [..] ... > http://lists.mech.kuleuven.be/mailman/listinfo/orocos-users >