By using the approach, the original optimal control problem of bilinear systems with disturbances is transformed into a sequence of nonhomogeneous linear two-point boundary value (TPBV) problems.