Abstract: The basic principle of correlation velocimetry sonar is "waveform invariance", which is described as follows: when a carrier transmits signals with a certain time interval, the distance between the two elements with the greatest correlation between the received echoes is proportional to the carrier's moving speed. The correlation velocimetry algorithm firstly constructs the data correlation matrix through the echo data, then uses the velocity estimation algorithm to fit the theoretical correlation model with the correlation matrix constructed by the echo data to get the optimal parameter information, and then calculates the carrier velocity information. In order to simplify the process of solving optimal parameters, a parameter fitting algorithm based on matrix derivation is proposed in this paper. In this method, the target parameters in the space-time correlation function model of bottom reverberation are separated, and then the function is expressed in matrix form. Finally, by matrix QR decomposition and local least squares, the analytical solution formula of the target parameters is obtained. The feasibility of this scheme as a velocity solution method is verified by simulation and on-lake test data, and it has the advantages of low computational complexity.