This paper presents a generalization of a technique developed previously for optimization of stochastic saturating systems. The generalization includes: (1) extension of the technique to systems with multiple control functions, (2) transformation of the optimization problem with constraints to the solution of a set of non-linear algebraic equations, and (3) a unified approach for different types of non-linear control functions. Also, a computation algorithm for solving the set of non-linear equations is proposed in this paper. Numerical examples are given for discussion on the questions of convergence and existence of a solution.