forwardKinematics for Delta printers

This commit is contained in:
AnHardt 2016-06-07 01:44:14 +02:00 committed by Scott Lahteine
parent 0b2c608d2f
commit 499e404fbf

View file

@ -7780,6 +7780,76 @@ void clamp_to_software_endstops(float target[3]) {
return abs(distance - delta[TOWER_3]); return abs(distance - delta[TOWER_3]);
} }
float cartesian[3]; // result
void forwardKinematics(float z1, float z2, float z3) {
//As discussed in Wikipedia "Trilateration"
//we are establishing a new coordinate
//system in the plane of the three carriage points.
//This system will have the origin at tower1 and
//tower2 is on the x axis. tower3 is in the X-Y
//plane with a Z component of zero. We will define unit
//vectors in this coordinate system in our original
//coordinate system. Then when we calculate the
//Xnew, Ynew and Znew values, we can translate back into
//the original system by moving along those unit vectors
//by the corresponding values.
// https://en.wikipedia.org/wiki/Trilateration
// Variable names matched to Marlin, c-version
// and avoiding a vector library
// by Andreas Hardtung 2016-06-7
// based on a Java function from
// "Delta Robot Kinematics by Steve Graves" V3
// Result is in cartesian[].
//Create a vector in old coords along x axis of new coord
float p12[3] = { delta_tower2_x - delta_tower1_x, delta_tower2_y - delta_tower1_y, z2 - z1 };
//Get the Magnitude of vector.
float d = sqrt( p12[0]*p12[0] + p12[1]*p12[1] + p12[2]*p12[2] );
//Create unit vector by dividing by magnitude.
float ex[3] = { p12[0]/d, p12[1]/d, p12[2]/d };
//Now find vector from the origin of the new system to the third point.
float p13[3] = { delta_tower3_x - delta_tower1_x, delta_tower3_y - delta_tower1_y, z3 - z1 };
//Now use dot product to find the component of this vector on the X axis.
float i = ex[0]*p13[0] + ex[1]*p13[1] + ex[2]*p13[2];
//Now create a vector along the x axis that represents the x component of p13.
float iex[3] = { ex[0]*i, ex[1]*i, ex[2]*i };
//Now subtract the X component away from the original vector leaving only the Y component. We use the
//variable that will be the unit vector after we scale it.
float ey[3] = { p13[0] - iex[0], p13[1] - iex[1], p13[2] - iex[2]};
//The magnitude of Y component
float j = sqrt(sq(ey[0]) + sq(ey[1]) + sq(ey[2]));
//Now make vector a unit vector
ey[0] /= j; ey[1] /= j; ey[2] /= j;
//The cross product of the unit x and y is the unit z
//float[] ez = vectorCrossProd(ex, ey);
float ez[3] = { ex[1]*ey[2] - ex[2]*ey[1], ex[2]*ey[0] - ex[0]*ey[2], ex[0]*ey[1] - ex[1]*ey[0] };
//Now we have the d, i and j values defined in Wikipedia.
//We can plug them into the equations defined in
//Wikipedia for Xnew, Ynew and Znew
float Xnew = (delta_diagonal_rod_2_tower_1 - delta_diagonal_rod_2_tower_2 + d*d)/(d*2);
float Ynew = ((delta_diagonal_rod_2_tower_1 - delta_diagonal_rod_2_tower_3 + i*i + j*j)/2 - i*Xnew) /j;
float Znew = sqrt(delta_diagonal_rod_2_tower_1 - Xnew*Xnew - Ynew*Ynew);
//Now we can start from the origin in the old coords and
//add vectors in the old coords that represent the
//Xnew, Ynew and Znew to find the point in the old system
cartesian[X_AXIS] = delta_tower1_x + ex[0]*Xnew + ey[0]*Ynew - ez[0]*Znew;
cartesian[Y_AXIS] = delta_tower1_y + ex[1]*Xnew + ey[1]*Ynew - ez[1]*Znew;
cartesian[Z_AXIS] = z1 + ex[2]*Xnew + ey[2]*Ynew - ez[2]*Znew;
};
#if ENABLED(AUTO_BED_LEVELING_FEATURE) #if ENABLED(AUTO_BED_LEVELING_FEATURE)
// Adjust print surface height by linear interpolation over the bed_level array. // Adjust print surface height by linear interpolation over the bed_level array.