In order to solve the real-time problem of nonlinear filter tracking using dynamic acoustic array in colored noise environment, an interacting multiple model target tracking method based on the improved unscented particle filter(IMM-IUPF) is proposed.In the method, an improved Sigma point selection strategy based on a minimal skew simplex unscented transform and a scaled unscented transform is adopted, which reduces the computational complexity of Sigma point selection in the standard unscented particle filter(UPF).Then the improved UPF is combined with the interacting multiple model(IMM).In addition, a self-adaptive gain modification coefficient is defined for solving the low-accuracy problem caused by the Sigma point reduction.Finally, the validity and superiority of the proposed algorithm is verified in comparison with traditional the IMM-PF and recent IMM-UPF.