[ODE] [Fwd: Re: Inverse Kinematics]

Jun Li junjy at dcs.qmul.ac.uk
Mon Mar 27 03:14:44 MST 2006



---------------------------- Original Message ----------------------------
Subject: Re: [ODE] Inverse Kinematics
From:    "Jun Li" <junjy at dcs.qmul.ac.uk>
Date:    Mon, March 27, 2006 11:14 am
To:      "Guillermo Andrade B." <guille.andrade at free.fr>
--------------------------------------------------------------------------

Hi,
I have the similar problem, And I don't want to adjust the parameter.

Thus I used AMotor, control the desired velocity, and observe
their output torque/force.

May I get 'realistic' torque/force by this way?
If not, why?

And I have an implementation problem either,
the same code doesn't work for cygwin, but on Mandrake Linux,
it works well.
By removing suspicious code gradually, I find the problem is
the amotor. If I remove amotor control, the simulation does
the exactly same thing on two platform.

Is there known problem about amotor on different platform?

Thanks!


> Hello
>
> You must control speed too:
>
> forceVector = scale*(goal_location_vector - hand_location_vector)
> 	      + scale2*(goal_speed_vector - hand_speed_vector)
> addForce(forceVector)
>
> But in some cases, when arm is near singular configurations, scale and
> scale2 must be changed in the aim to preserve stability. See keys :
> Jacobian, Gain Matrix, stability
>
>
> best regards,
>
> Guillermo Andrade
>
>
> Bernie Roehl a écrit :
>> Hi everyone
>>
>> I'm doing a robot arm simulation, and I want to have the robot simply
>> reach toward a goal point and follow the goal as it moves.  I could use
>> IK, but I'm thinking there's probably a way to do it using ODE.
>>
>> I'm implemented something similar to the double-pendulum example, but
>> with a force that moves the second pendulum towards the goal point.
>> Essentially I'm doing this:
>>
>> forceVector = scale*(goal_location_vector - hand_location_vector)
>> addForce(forceVector)
>>
>> Of course, what I get from that is just a pendulum -- the hand
>> accelerates towards the goal point, and travels past it, before the
>> force pulls it back towards the goal again, and so on.
>>
>> Anyone have any suggestions on a good way to do this?  Ideally, I'd like
>> to have the hand accelerate and then decelerate so the final velocity is
>> zero when it reaches its goal.
>>
>> Thanks in advance for any suggestions.
>>
>>
>>
>
> _______________________________________________
> ODE mailing list
> ODE at q12.org
> http://q12.org/mailman/listinfo/ode
>


-- 
Jun Li PhD Student
CS 319 Department of Computer Science
Queen Mary, University of London
Mile End London E1 4NS
+44 (0)20 7882 5230



-- 
Jun Li PhD Student
CS 319 Department of Computer Science
Queen Mary, University of London
Mile End London E1 4NS
+44 (0)20 7882 5230




More information about the ODE mailing list