For solar array system, traditional modeling theories usually adopt the Cartesian coordinates to establish the dynamic equation of the system. The order of this equation is often too high and inconvenient to design control laws and the Cartesian coordinates are difficult to be measured in practice. This paper presents a modeling method for solar array system. In this method, joint coordinates of the solar panels, yoke and the spacecraft main-body are defined as the generalized variables. In this way, the dynamic equation derived here possesses lower order than those derived by other methods used before, which facilitates further control design. Moreover, the joint coordinates can be easily measured in practice. On the other hand, joint friction is an important issue to be considered. Joint friction could affect the stability and control precision of the system. Most existing studies on joint friction of solar array system are over-simplified into plane problems, and as a result, inaccuracy of friction force calculation output was incurred. In this paper, a three-dimensional revolute joint model is introduced, the calculation of joint friction is discussed in detail, and the relationship between ideal constraint force and Lagrange multipliers is derived. In addition, the control design of the solar array system is discussed and the fuzzy adaptive PD control method is used for controller design. At the end of this paper, the validity of the above studies is verified by the comparisons of numerical simulations with the ADAMS software.