In decision-making environment, due to the influence of external environment and market conditions, the profit can not be estimated accurately, in this case, interval number can well represent the possible range of return value. Firstly, considering the rationality of profit allocation, the residual vectors were obtained by the players participating in the alliance were equally distributed. Based on the square distance of interval values and the least square method, a quadratic programming model was constructed to minimize the sum of the square excess of interval values of all players in the maximum cooperative alliance, and the profit allocation value of each player was determined. Then, considering the influence of the player's weight on the profit allocation strategy, the corresponding quadratic programming model was constructed, and the analytical formula of the optimal solution was obtained, that is, the weighted interval number equally divided the residual value. Finally, taking the cold chain logistics enterprise alliance as an example, the proposed (weighted) interval number average residual value was applied to solve the actual cooperative income allocation strategy problem. The results show that the solutions meet the characteristics of cooperative game such as existence, uniqueness, effectiveness, individual rationality, super additivity, symmetry, etc. The profit allocation strategies are fair and reasonable.