Font Size: a A A

The Optimization Analysis Of Mixture Uniform Design Based On Genetic Algorithm

Posted on:2012-11-01Degree:MasterType:Thesis
Country:ChinaCandidate:J S ZhouFull Text:PDF
GTID:2154330332496595Subject:Epidemiology and Health Statistics
Abstract/Summary:PDF Full Text Request
Objective: Research principle and application of the mixture uniformdesign. Explore the Mixture of single-objective genetic algorithm optimaluniform design formula, the ratio of the optimization results. Overcome thetraditional methods of subjectivity and lack of local optimization toimprove the accuracy of optimization algorithm. Provide Scientific andrational approach for Medicine mixture uniform design for the optimizationproblem. Expand the scope of application software base on the Geneticalgorithm platforms to, to provide practical tools.Methods: Research the principle of mixture uniform design. Arrange thetest Reasonable. Construction Scheffe polynomial model of mixtureuniform design which Based on the Constraints of and other conditionsattached under the common constraint in the mixture. Using the Geneticalgorithm platforms v1.0 software to evaluated the effect of geneticalgorithm optimization. Explore the Genetic algorithm with the uniformmixing formula, the ratio of the optimization results. It compared with theresults of Mathematical derivation and three-dimensional map and contourmap. For paracetamol and oral disintegrating tablets prescriptionsinomenine the ratio of liposome membrane problem using geneticalgorithm constrained single objective optimization, and compared with thetraditional optimization methods.Results: (1) great value 10.136973 test functions, and not within theconstraints, constraints in the context of its contour lines only solution;genetic algorithm to search within the constraints of the objective functionvalue of the average level of 10.136 , variability of small, high precision,and the contour map visually see the results are consistent, the solution isconstrained within the contour line of the solution;(2) Obtained by genetic algorithm in Example 1 of paracetamol orallydisintegrating tablet hardness average of 91.180, higher than the hardness of the original forecast of 87.3 in the literature, increased 6.72%; geneticalgorithm to decomposition of the target groups Were also larger than anyone test the hardness value. If you choose the first 13 search results as theoptimal formulation conditions, the oral disintegrating tablet hardness of93.174, corresponding to the ratio of the prescription:Lactose takesPercentage of 30.1%, microcrystalline cellulose takes percentage of 60.3%, crosslinked sodium carboxymethyl cellulose takes the Percentage of 9.6%(3) obtained by genetic algorithm example 2 sinomenine prescriptionencapsulation efficiency of liposome membrane average of 80.6%, higherthan originally forecast in the literature entrapment efficiency 75.36%,7.61% increase; genetic algorithm to The decomposition of the targetgroups are larger than any one test the prediction encapsulation efficiency.In this case, the highest encapsulation efficiency was predicted, thedistribution ratio of each group under the conditions were:Sinomenine0.1%, 75.6% soybean lecithin, cholesterol, 23.9%, vitamin E 0.4%.Conclusiononclusion: After the study results suggest that the proportion of geneticalgorithm components reasonably good results can be used for single-targetmixture uniform design optimization, and can guide the practicalapplication.
Keywords/Search Tags:Mixture uniform design, Genetic algorithm, Optimizationof single Objective, Constraints
PDF Full Text Request
Related items