The invention relates to a multi-moving-target positioning method based on an improved TDOA/FDOA algorithm, and mainly solves the problems that the adaptive capacity of noise errors is poor and the positioning precision is limited when a traditional algorithm is used for positioning a plurality of dynamic targets. The method comprises the following steps: firstly, setting K moving targets which do not coincide with one another in a three-dimensional space, deploying M sensors, and acquiring a measurement vector of target information; then constructing a TDOA/FDOA positioning equation set about the position information of the target and the sensor so as to estimate the speed and the position of the target; constructing a global constraint condition according to a relationship between the targets and additional variables, introducing a Lagrange multiplier technology, and adopting a quasi-Newton BFGS iterative formula to solve a positioning equation. According to the method, constraint conditions are comprehensively and systematically considered, and the Lagrange multiplier technology and the quasi-Newton BFGS iterative formula are introduced, so that calculation of a Hessian matrix is avoided, the calculation amount is greatly reduced, and the positioning precision is improved.