The relation between the direction cosine matrix and Euler angles is:
R=
cosdcosy singsindcosy—cos¢siny cos¢sin@cosy+singsiny/
cosOsiny singsindsiny+cosdcosy cos¢sin @siny—sin pcos
-sing singcosd cos pcos
Equation 1 and equation 2 expresse how to rotate a vector measured in the
frame of reference of the plane to the frame of reference of the ground.
Equation 1 is expressed in terms of direction cosines. Equation 2 is
expressed in terms of Euler angles.
Eqn. 2
In equation 1, each component of the vector in the ground frame is equal to
the dot product of the corresponding row of the rotation matrix with the vector
in the plane frame. Nine multiplies and six additions are required to compute
the rotation. Equation 3 is a restatement of equation 1, with the matrix
multiplication expanded in terms of the elements of the vectors and the
matrix.
Qn = at Qre Hy Qry Hc Qrz
6, =F Qretly QoQ Eqn. 3
Gs = Fx Qoehy Qo hs Qrs
Note that the R matrix is not necessarily symmetric. The three columns of
the R matrix are the transformations of the three axis vectors of the plane to.
the ground frame of reference. The three rows of the R matrix are the
transformations of the three axis vectors of the ground coordinate system to
the plane frame of reference. The R matrix contains all the information
needed to express the orientation of the plane with respect to the ground. The
R matrix is also called the direction cosine matrix, because each entry is the
cosine of the angle between an axis of the plane and an axis on the ground.
Although it would appear that there are 9 independent parameters in the R
matrix, there are really only 3 independent ones, because of the six so-called
orthogonality (also known as normalization) conditions: the three column
vectors are mutually perpendicular and the magnitude of each column vector
is equal to one.
The transpose of any matrix, and the rotation matrix in particular, indicated
as R', is formed by interchanging rows and columns. In general, the inverse
of a square matrix, if it exists, is indicated as R~’. The inverse of a matrix
times the matrix produces the identity matrix. (The identity matrix has all ones
on the diagonal, and all zeros everywhere else. Multiplying any matrix by the
identity matrix leaves it unchanged. In the case of rotation matrices, it turns
out that the transpose of the R matrix is equal to its inverse:
DCM 10 Draft: §/17/2009Eqn. 4
Q, =R"Q, =R'Q,
The reason that the inverse of the rotation matrix is equal to its transpose is
because of the symmetry of the situation. The elements of the rotation matrix
are the cosines between pairs of axis, one in the plane frame, and one in the
ground frame. The inverse situation is equivalent to exchanging the roles of
the ground and plane frame of reference, which is the same as interchanging
rows and columns, which is the same as the transpose.
Also, the fact that the inverse is equal to the transpose is consistent with the
orthogonality conditions, which can be expressed in matrix notation as:
100
=j0 10 Eqn. 5
001
Equation 5 can be used to prove that the inverse of R is R transpose, by
multiplying the equation by the inverse of R, or by the inverse of R transpose.
A very useful property of the rotation matrix is that we can compose
rotations. We can multiply several rotations matrices together, and get a
rotation matrix that is equivalent to applying all of the rotations in succession.
We have to be careful to apply the rotations in succession on the left side of
what we already have. For example, if we have three rotation matrices, from
orientation A to orientation B, from B to C, and from C to D, we can compute
the rotation matrix that will go from orientation A to orientation D according to:
R'=R'R
fotation matrix from Bto C Eqn. 6
rotation matrix from C to D
Ryq = rotation matrix from A to D
The reason that we have to be careful about the sequence of operations
when multiplying rotations matrices is that matrix multiplication is NOT
commutative. That is, the order of matrix multiplication matters very much...
This is consistent with rotations, which are not commutative either. For
example, consider what happens if a plane pitches around its own pitch and
roll axes by 90 degrees each. The order very much matters. Suppose that is
pitches up by 90 degrees, followed by a roll of 90 degrees. At that point the
plane will be traveling vertically. However, if it rolls first, and then banks, it will
be traveling in the horizontal plane.
DCM nN Draft: §/17/2009Finally, there is a useful identity that applies to matrices in general, and to
rotation matrices in particular. The transpose of the product of two matrices is
equal to the product of the transposes of the matrices, with the two matrices
swapped:
(AB) =B'A Eqn.7
‘A,Barematrices
Vector dot and cross products
Two very useful vector products that we will use in computing DCM and in
using its elements for navigation and control are the dot product and the cross
product. The dot product of two vectors A and B, is a scalar computed by
performing a matrix multiplication of a A as a row vector with B as a column
vector producing:
B
A-B=A'B=[4, 4, A]]B,|=
8,
It turns out that the vector dot product produces a result that is equal to the
product of the magnitudes of the two vectors, times the cosine of the angle
between them:
1,B,+A,B,+AB, Eqn. 8
A-B=|A[B|-cos@,.) Eqn. 9
We note that the dot product is commutative: A-B=B-A
The cross product of two vectors A and B, is a vector whose components
are computed by:
(AxB), =4,8,-AB,
(AxB),=4.8,-AB, Eqn. 10
(AxB), =4,B,-A,B,
The cross product is perpendicular to both of its vector factors and its
magnitude is proportional to the magnitudes of the vectors times the sine of
the angle between them:
DCM 12 Draft: §/17/2009