Derek,

Try this:

http://www.dynomotion.com/Help/LibrariesFlowDiagram.PNG

But yes, it should be almost as simple as swapping your kinematics for the “3Rod” equations.  You will need MS Visual Studio 2008 Standard edition to compile and make changes.  Place any file called Kinematics.txt into the \KMotion\Data directory to cause the CKinematics3Rod class to be used by the system instead of the standard base class CKinematics.

In the case of the 3Rod (3 linear actuators) the inverse kinematics are very simple to compute.  See the function shown below from file Kinematics3Rod.cpp:

int CKinematics3Rod::TransformCADtoActuators(double x, double y, double z, double a, double b, double c, double *Acts)

{

      // find lengths of each actuator

      GeoCorrect(x,y,z,&x,&y,&z);

      double r0 = sqrt(sqr(x-Act0Center.x) + sqr(y-Act0Center.y) + sqr(z-Act0Center.z)) - Act0Off;

      double r1 = sqrt(sqr(x-Act1Center.x) + sqr(y-Act1Center.y) + sqr(z-Act1Center.z)) - Act1Off;

      double r2 = sqrt(sqr(x-Act2Center.x) + sqr(y-Act2Center.y) + sqr(z-Act2Center.z)) - Act2Off;

      Acts[0] = r0*m_MotionParams.CountsPerInchX;

      Acts[1] = r1*m_MotionParams.CountsPerInchY;

      Acts[2] = r2*m_MotionParams.CountsPerInchZ;

      Acts[3] = a*m_MotionParams.CountsPerInchA;

      Acts[4] = b*m_MotionParams.CountsPerInchB;

      Acts[5] = c*m_MotionParams.CountsPerInchC;

      return 0;

}

The forward Kinematics are more difficult to compute so they are automatically computed!  (numerically using the inverse Kinematics in the function):

int CKinematics3Rod::TransformActuatorstoCAD(double *Acts, double *xr, double *yr, double *zr, double *ar, double *br, double *cr)

There is a parameter that will set the maximum length along a path where a linear relationship between CAD and Actuator space may be assumed.

      m_MotionParams.MaxLinearLength = 0.25;  // limit the segment lengs for nonlinear systems

I think a powerful feature of our Trajectory planner is to handle planning through many tiny segments each with different velocity and acceleration constraints.

Regards,

TK