球坐标变换、主成分二次推断函数及改进非劣分类遗传算法在不含“0”缓释剂混料组分配比优化的效果研究

The Effect of Optimization of Mixture Ratio without“0”Sustained Release Agent based on Hyperspherical Transformation,Principal Component Quadratic Inference Function and NSGA-Ⅱ

目的针对不含“0”混料组分配比中重复测量数据及不同时点累计释放度的多目标优化,研究基于球坐标变换的主成分二次推断函数和改进非劣分类遗传算法优化混料处方配比的效果。方法对尼莫地平缓释片不含“0”混料的处方配比的试验数据进行处理,使用球坐标变换方法去除定和约束,主成分二次推断函数建立模型解决组分的共线性及重复测量数据的相关性,根据药典对各时点累积释放度的要求,采用改进非劣分类遗传算法进行多目标优化,寻找最优的配比方案。结果当HPMC、乳糖、海藻酸钠的比例分别为26.49%、64.64%、8.87%时,3h、6h、9h、12h的累积释放度分别为21.21%、50.66%、77.60%、99.60%,均在处方筛选范围。比通过建立Scheffé多项式模型寻找到5个最优方案更好。结论针对不含“0”混料组分缓释制剂重复测量数据处方配比的多目标优化,采用球坐标变换的主成分二次推断函数和改进非劣分类遗传算法优化混料处方配比的效果满意,且研究者可在Pareto非劣解集中选择可行、合理的混料成分配比方案。

Objective Aiming at the multi-objective optimization of repeated measurement data without“0”in the mixture ratio,the effect of the quadratic inference function of principal component based on hyperspherical transformation and the non-dominated sorting in genetic algorithm-Ⅱto optimize the mixture prescription ratio was studied.Methods To nimodipine matrix tablet excluding“0”of the blending ratio of the prescription of the test data,using hyperspherical transformation to eliminate and constraints,the quadratic inference function of principal component to solve the components of collinearity and the correlation of repeated measurement data,according to the requirement of the pharmacopoeia for releasing degrees each point accumulation,with the NSGA-Ⅱfor multi-objective optimization,search for the optimal proportion plan.Results When the proportions of HPMC,lactose,and sodium alginate were 26.49%,64.64%,and 8.87%,respectively,the cumulative release degrees at 3h,6h,9h,and 12h were 21.21%,50.66%,77.60%,and 99.60%,respectively,which were within the scope of prescription screening.It is better than the five optimal schemes by establishing Scheffépolynomial model.Conclusion Aiming at the multi-objective optimization of the prescription of repeated measurement data of the sustained release preparation without“0”mixed components,the quadratic inference function of principal component of hyperspherical transformation and the NSGA-Ⅱwere used to optimize the mixture prescription ratio satisfactorily.Moreover,the researcher could select a feasible and reasonable mixture composition scheme in Pareto non-inferior solution concentration.