Formulation of the problem. Classical optimal Bayesian phase lock loop (PLL) algorithms require a priori knowledge of the phase dynamics parameters and signal-to-noiz ratios of the received signals. In practice, these parameters vary significantly and, as a rule, are unknown. Taking into account the fact that the operation of such algorithms under conditions different from those a priori specified is not reliable according to the criterion of minimum variance error. Purpose of the study. Develop a phase-locked loop system for the Global Navigation Satellite System (GNSS) signal that adapts to phase dynamics and signal-to-noise ratio to maintain phase tracking over the widest possible range of operating conditions. Results. A multi-channel adaptive phase-locked loop system (MAPLL system) has been developed. Practical significance. The developed system is capable of processing a jump in the signal-to-noise ratio from 50 to 9 dBHz and back without losing phase tracking (under conditions of low dynamics), maintaining phase tracking during abrupt transitions of dynamics between low (due only to the dynamics of the reference oscillator) and high (sinusoidal acceleration 10g and sinusoidal jerk 10 g/s) at a signal-to-noise ratio of 24 dBHz. Thus, in real-world conditions where object dynamics and S/N ratios of received signals change in unpredictable ways, MAPLL system maintains phase tracking over a much wider range of conditions than non-adaptive PLL.