You are on page 1of 2

Matrix Representation of an Incremental Rotation

William Premerlani In the documentation of the direction-cosine-matrix algorithm, there is a step in the derivation of the theory that is difficult for many readers. It is the step from equation 15 to equation 17 in the document. The two equations in question are shown here as equations 1 and 2. Equation 1 expresses the incremental change in a coordinate axis, in terms of an increment in the rotation vector:

rearth (t + dt ) = rearth (t ) + rearth (t ) d(t ) d(t ) = (t )dt


Equation 2 expresses 3 instances of equation 1 in terms of the multiplication of two 3 by 3 matrices: 1 R (t + dt ) = R (t ) d z d y d x = x dt d y = y dt d z = z dt d z 1 d x d y d x 1

Eqn. 1

Eqn. 2

The equivalence of equation 1 and 2 can be shown by first expressing equation 2 in mixed vector-matrix notation:

r1 (t + dt ) r1 (t ) 1 r (t + dt ) = r (t ) d z 2 2 r3 (t + dt ) r3 (t ) d y r1 = first row of R r2 = second row of R r3 = third row of R

d z 1 d x

d y d x 1

Eqn. 3

In equation 3, the direction-cosine matrix is represented as three row vectors with three elements each, instead of a single matrix with 9 elements. Each row vector represents one earth axis as seen in the body frame. The subscripts 1, 2, and 3 refer to the x, y, and z axis of the earth frame. (I have used 1, 2, and 3 to avoid confusion with the x, y, and z components of the vectors.) We see that the transformation operation is the same form for each axis, so we can consider one axis at a time and drop the subscript. The result is equation 4, where it is to be understood that equation 4 should be applied 3 times for each row of the direction cosine matrix. (As an aside, the basis for DCM-light can be understood from equation 4. Each row of the direction DCM rotation 1 February 1, 2010

cosine matrix is updated using all elements of the gyro rate vector, but only the matrix elements of that row. Therefore, if you are interested in only one of the rows of the direction cosine matrix, it can be updated using information from only that row, so you can ignore the other two rows.)

1 d z d y 1 d x r (t + dt ) = r (t ) d z Eqn. 4 d y d x 1 Equation 4 can be expanded in terms of the x, y, and z components of the vector:

1 d z d y rx (t + dt ) ry (t + dt ) rz (t + dt ) = rx (t ) ry (t ) rz (t ) d z 1 d x Eqn. 5 d y d x 1 After performing the vector-matrix multiplication in equation 5 and expressing the result as three equations, we find:

] [

rx (t + dt ) = rx (t ) + ry (t ) d z rz (t ) d y ry (t + dt ) = ry (t ) + rz (t ) d x rx (t ) d z rz (t + dt ) = rz (t ) + rx (t ) d y ry (t ) d x
The second and third terms in each equation are recognizable as a cross product. Therefore, equation 6 can be re-expressed in vector form as:
r (t + dt ) = r (t ) + r (t ) d(t )

Eqn. 6

Eqn. 7

Therefore, equations 1 and 2 are equivalent.

DCM rotation

February 1, 2010

You might also like