Selecting cooperative innovation is the optimum game development strategy, with which an enterprise realize its strategic objectives and obtains the technological breakthrough. Cooperative partner's assessment and selection is the primary link, it decides the efficiency and effect of cooperative innovation to a certain extent. In order to be scientifically effective and easy to realize the choice of the cooperative innovation partner, this paper combines PSO and ameliorated TOPSIS Methodtogether and proposes a new method for solving multiple attribute decision limited scheme -based on PSO fixed weight and ameliorated TOPSIS method. To establish a nonlinear programming problem about weight, according to the index weight of TOPSIS method, through the optimal and with the index for standard minimum sum object distance,present its mathematical programming model, and use PSO to determine the weight of the index; on this basis, in order to overcome the shortage of the traditional TOPSIS method, based on set pair analysis relative degree, set the ideal point and negative ideal point as set of oppositionabout certain and uncertain system, in examining the relative degree of alternatives and ideal point or negative ideal point, fully considering the existence of the opposite set,establishes comprehensive evaluation model of ameliorated TOPSIS method based on relative degrees, and applies it in the empirical analysis on the selection of the new technology innovative cooperation partner for H automobile group, and verifies the rationality and validity of the method.