It is convenient to keep equations in the form expression == 0 and furthermore functions dealing with equations will typically understand a given expression as an equation of that form. Hence, you can avoid using == in favor of using the difference between the lhs and rhs:

eq_a = diff(r_a, t) - v0*(r_b - r_a)/(r_b - r_a).norm()

If you still want to get explicit equations with == from the above vector - here is a way:

eq_a_eq = [eq==0 for eq in eq_a]

Well... vector/matrix equations are a tad awkward :

sage: from _operator import eq
sage: vector(map(eq, diff(r_a, t), v0*(r_b - r_a)/(r_b - r_a).norm()))
(diff(x_a(t), t) == -v0*(x_a(t) - x_b(t))/sqrt(abs(-x_a(t) + x_b(t))^2 + abs(-y_a(t) + y_b(t))^2), diff(y_a(t), t) == -v0*(y_a(t) - y_b(t))/sqrt(abs(-x_a(t) + x_b(t))^2 + abs(-y_a(t) + y_b(t))^2))

$$\left(\frac{\partial}{\partial t}x_{a}\left(t\right) = -\frac{v_{0} {\left(x_{a}\left(t\right) - x_{b}\left(t\right)\right)}}{\sqrt{{\left| -x_{a}\left(t\right) + x_{b}\left(t\right) \right|}^{2} + {\left| -y_{a}\left(t\right) + y_{b}\left(t\right) \right|}^{2}}},\,\frac{\partial}{\partial t}y_{a}\left(t\right) = -\frac{v_{0} {\left(y_{a}\left(t\right) - y_{b}\left(t\right)\right)}}{\sqrt{{\left| -x_{a}\left(t\right) + x_{b}\left(t\right) \right|}^{2} + {\left| -y_{a}\left(t\right) + y_{b}\left(t\right) \right|}^{2}}}\right)$$

HTH,