Unbalanced water flooding can easily occur due to the influence of reservoir heterogeneity and development measures. Based on the statistical method，the Christiansen uniformity coefficient is used to quantitatively evaluate the balanced water flooding degree during the development of multi-layer reservoirs. A mathematical model for optimizing multi-layer injection and production parameters is established to maximize balanced water flooding，and the covariance matrix adaptive intelligent evolution algorithm is used to solve the problem. A collaborative optimization method of injection and production parameters for multi-layer reservoirs based on water flooding is developed，and its reliability is verified by an example. The new method is applied to the typical well group of Shengtuo T142 reservoir in Shengli Oilfield. By comparing the results before and after the optimization of intelligent collaborative stratified injection and production parameter，it is found that the Christiansen uniformity coefficient by the original scheme is improved by 3.49%. The cumulative oil production is increased by 1.024 × 10⁴m³ ，and the water cut is decreased by1.34%. The results show that this method can accurately control the production and injection allocation between layers，effectively improve the balanced water flooding between layers and wells，and realize the accurate matching between injection and production parameters and remaining oil distribution.