Based on dynamic particle swarm optimization algorithm (DPSO) and transfer matrix method,a novel method is presented to reconstruct the axial non uniform strain profile distribution of fiber Bragg grating (FBG).The strain value of sampling points from fiber gratings axial are presented in the form of particle,and the optimized strain value are obtained by the particles′ searching in the solution space according to the birds behaviors.The inertia weight of this algorithm will be changed dynamically according to the distance between different particle and the global optimal particle in current population;as a result,the speed of convergence to optimal position is faster.DPSO algorithm and quantum behavior particle swarm optimization algorithm (QPSO) are used for reconstructing four types of strain profiles,including linear,quadratic,sine and discontinuous profiles. The simulations demonstrate that DPSO algorithm can effectively reconstruct the axial non uniform strain profile distribution along the whole FBG,and its iterative speed and accuracy are faster than QPSO method.