Parametric optimization problems for hydromechanical systems with continuous not everywhere differentiable multiextremal criteria both in scalar and in vector statement are considered. Global solutions for individual criterion are determined using new hybrid algorithms that combine the Metropolis algorithm and deterministic methods for local search. The vector optimization algorithms generate a set of non-dominated solutions which constitute Pareto-optimal front approximations. The proposed hybrid algorithms can be applicable for computational diagnostics systems, design optimization, complex dynamic systems control.