One of the important things in the dynamic measurements theory is the task of dynamically distorted signal restoration. The existing algorithms demand a bigger amount of machine time. In the project the numerical algorithm of the optimum dynamic measurement problem solution with the use of parallelization and linear splines is presented. A distinctive feature of the algorithm is that one was able to use the parallelization procedure by means of consideration of the optimum dynamic measurement problem as a mixed optimum control problem. Previously, for the solution of the optimum dynamic measurement the problems of rigid and optimum control were used. We emphasize that the functionality of the quality of the mixed optimum dynamic measurement problem preserves the basic principle of dynamically distorted signal restoration — achievement of proximity between virtual and real observations. The algorithm is adapted for effective usage on the computing system with parallel architecture or the multiprocessor computing system. Parallelization as well as the use of linear splines at one of the algorithm stages will allow to reduce the amount of machine time necessary for calculations.