以氮、磷、钾、密度及果枝数五项农艺措施为决策变量,采取五因素二次回归旋转设计方案,经田间试验后进行计算机处理,建立了相应的数学模型。本文对因素效应和不同产量等级筛选出的优化方案进行了分析。结果表明,本棉区以密度和钾肥对产量影响最显著。≥75公斤/亩产量的优化方案为N10.45～11.05公斤/亩,P_2O_53.92～4.10公斤/亩,K_2O10.65～11.20公斤/亩,密度3410～3510株/亩,果枝数16.06～16.30台/株。 Taking five agronomic measures of nitrogen, phosphorus, potassium, density and fruiting branch number as decision variables, a five-factor quadratic regressive rotation design was adopted. After the field experiment, the computer processing was carried out to establish the corresponding mathematical model. In this paper, the optimization effect of factors and different output levels were analyzed. The results showed that the cotton area density and potassium fertilizer on the yield the most significant. The optimal plan of the yield of ≥ 75 kg / mu is N10.45 ~ 11.05 kg / mu, P_2O_53.92 ~ 4.10 kg / mu, K_2O10.65 ~ 11.20 kg / mu, density of 3410 ~ 3510 plants / mu and fruiting branches of 16.06 ~ 16.30 Taiwan / strains.