Artificial Immune Network-Based Pharmacokinetic Study

Artificial immune network model is a natural computational model to solveengineering and scientific problems and it provides a way for information processingand computing. At present, algorithms based on artificial immune network modelmainly include the following: clonal selection algorithm, negative selection algorithm,immune learning algorithm, hybrid immune algorithm, etc. In recent years, research andimprovement of artificial immune network algorithm has become a hot research topic inthe field of artificial immune. The new direction in recent years is to apply artificialimmune network algorithm to the pharmacokinetics parameters optimization research.The pharmacokinetics is a subject focuses on discipline of in vivo dynamic variation ofthe drug over tome, and for all current method of parameter optimization inpharmacokinetics there is one existing shortcoming of no precise convergence forparameter optimization.The principal targets in this thesis takes optimization as the lead, and it furtherdiscusses the artificial immune network algorithm--opt-ainet, and applies it intomulti-objective optimization simulation experiment to test its algorithm performance.And then it proposes the artificial immune network hybrid algorithm--NetHJ, andapplies this algorithm in the test of pharmacokinetic model parameters optimization toanalyze the algorithm’s optimization ability. At last, by applying the NetHJ algorithm todata analysis example it further verified the effectiveness of algorithm in practical use.The research results conducted in this article are mainly as following:(1) This thesis carried out the research on artificial immune network algorithm--opt-ainet, and tests its optimization ability while taking other algorithms ascomparison.(2) Targeted on defects of the artificial immune network algorithm’s poor accuracylocal search, combined with HJPSM’s fine performance of local search accuracy thisthesis proposes a new artificial immune network hybrid algorithm—NetHJ.(3) It studies the artificial immune network algorithm--NetHJ and extends itsapplication fields of applying it into the optimization problem of compartment model pharmacokinetic parameters and nonlinear pharmacokinetic parameters optimization.And lastly it makes a comparison of optimization performance among commonly usedpharmacokinetic parameter optimization method to test practical application ability ofNetHJ algorithm in pharmacokinetic parameters optimization.(4) It furthers on the application and development of artificial immune networkalgorithm--NetHJ hybrid algorithm, and applies it to data analysis cases. It also makesprocess design for data analysis algorithm, develops with vc++language and realizesits operation. It further verified the practical application ability of artificial immunenetwork hybrid algorithm--NetHJ algorithm.