基于改进PSO的高分子材料降解模型参数优化

为提高可降解高分子材料降解模型仿真的准确程度,结合高分子材料降解的实际原则和所要考虑的各种因素,建立了适合优化的参数优化模型,并将粒子群优化算法(PSO)用于模型的求解.针对标准粒子群算法存在的一些不足,提出了一种改进的粒子群优化算法来求解最优值,改进的算法引入了动态自适应惯性权重和异步时变学习因子.采用5个标准测试函数对改进的粒子群算法进行了测试,并将算法应用于参数优化模型的求解.测试与试验结果表明:新算法有效地避免了过早陷入局部最优,提高了收敛速度和收敛精度,并且采用优化所得参数显著地提高了高分子材料降解模型仿真的精准度,有利于揭示降解机理的科学意义和指导实际医用器件的设计与生产.

Parameter optimization of polymer materials degradation model based on improved PSO algorithm

To improve simulation accuracy of biodegradable polymer materials degradation model,a mathematical optimization model based on the actual principles in degradation of polymer materials and various factors was proposed.The particle swarm optimization(PSO) was used to optimize model para-meters.To solve the weakness of standard particle swarm optimization algorithm,an improved algorithm was proposed to achieve optimization values.Dynamic adaptive inertia weight and asynchronous time-varying learning factors were introduced into the improved algorithm.The improved PSO algorithm was tested by five standard test functions to optimize the model parameters.The test and experiment results show that the proposed algorithm can effectively avoid being trapped in local minimum with increased accuracy and convergence rate.The simulation accuracy of biodegradable polymer materials degradation model can be significantly improved by optimized parameters.The algorithm is suitable to demonstrate the degradation mechanization and to guide the design and product of medical devices.