Optimization of the kinetic parameters with thermal analysis data is an essential issue. In this work, the conjugate direction particle swarm optimization (CDPSO) approach, as a global stochastic optimization algorithm that is suitable to high-dimension optimization problem, is first employed to estimate the kinetic parameters with DSC data. This algorithm combines the global optimization ability of the particle swarm optimization (PSO) and the ability to escape from the local extremum by the conjugate direction algorithm (CD) to find the globally optimal solutions. This algorithm does not require estimation of the initial values of the kinetic parameters. The validation of this method is verified by two cases: DCPO decomposition and CHP decomposition. By comparing the experimental and calculated heat flow results, the accuracy of the fitted kinetic parameters is verified. These two cases prove the effectiveness of CDPSO algorithm in the estimation of high-dimension kinetic parameters using DSC data.