In a multistatic sonar localization system, the sonar positions often include random errors that can seriously affect the target's localization accuracy. To tackle this problem, a moving target location algorithm based on time sum and Doppler frequency measurements is proposed. Firstly, the proposed algorithm reorganizes the nonlinear equations based on time sum and Doppler frequency localization mechanism into pseudo linear equations about target position, velocity and intermediate variables, and the initial moving target position and velocity are obtained by using weighted leastsquares estimates. Secondly, the correlation between the position, velocity and the intermediate variables is used to estimate the estimation biases of target position and velocity. Finally, an error correction technique is used to correct the initial target position and velocity estimates. The effectiveness of the proposed algorithm is theoretically analyzed under small measurement noise postulation and the numerical verification is performed by using Monte Carlo simulation.