The paper considers aspects of the development of algorithms for optimizing complex systems. The principles of constructing procedures for adjusting the parameters of variable gradient optimization algorithms are proposed. In any iterative algorithm, there are parameters that require their adjustment. To control and adjust parameters, - criteria are formed that determine the quality of adjustments. The problem of determining the best value of adjustment parameters belongs to the same class as the original optimization problem. During the operation of algorithms, their parameters are adapted to the original values. The paper provides recommendations for software implementation of probabilistic algorithms and construction of computational procedures for probabilistic computational experiments based on them.