This paper proposes a stochastic optimal technique based on Genetic Algorithm (GA) for the calculation of multiphase and multicomponent chemical equilibrium by minimization of Gibbs free energy. Three aspects are impr...This paper proposes a stochastic optimal technique based on Genetic Algorithm (GA) for the calculation of multiphase and multicomponent chemical equilibrium by minimization of Gibbs free energy. Three aspects are improved based on the drawbacks of the conventional GA.An alternative decimal encoding strategy is adopted to enhance the precision of calculation.A dynamic encoding method that can limit the bounds of optimized variables within their feasible regions is developed to cope with the complex constraints of the problem.Finally,sequential search technique is applied to improve GA to approach global optima.It is shown through the calculation of complex chemical systems,in which non-ideal,multireaction and multiphase coexistence are simultaneously involved,that the presented GA is general and efficient for the addressed problem.展开更多
为解决蒙特卡洛(Monte Carlo)方法在计算风险价值(Value at Risk,VaR)方面的缺陷,文章首先引入GARCH模型来刻画金融数据的波动聚集性,再引入马尔科夫链蒙特卡洛(Markov Chain Monte Carlo,MCMC)方法,来克服GARCH模型参数估计约束条件带...为解决蒙特卡洛(Monte Carlo)方法在计算风险价值(Value at Risk,VaR)方面的缺陷,文章首先引入GARCH模型来刻画金融数据的波动聚集性,再引入马尔科夫链蒙特卡洛(Markov Chain Monte Carlo,MCMC)方法,来克服GARCH模型参数估计约束条件带来的估计误差。通过对上证50指数的实证分析表明,引入MCMC方法可以提高模型的估计精确度。展开更多
基金partly supported by the China Postdoctoral Science Foundation(Grant No.2017M610156)the National Natural Science Foundation of China(Grant No.11501167)the Young Academic Leaders Project of Henan University of Science and Technology(Grant No.13490008)
文摘This paper proposes a stochastic optimal technique based on Genetic Algorithm (GA) for the calculation of multiphase and multicomponent chemical equilibrium by minimization of Gibbs free energy. Three aspects are improved based on the drawbacks of the conventional GA.An alternative decimal encoding strategy is adopted to enhance the precision of calculation.A dynamic encoding method that can limit the bounds of optimized variables within their feasible regions is developed to cope with the complex constraints of the problem.Finally,sequential search technique is applied to improve GA to approach global optima.It is shown through the calculation of complex chemical systems,in which non-ideal,multireaction and multiphase coexistence are simultaneously involved,that the presented GA is general and efficient for the addressed problem.
文摘为解决蒙特卡洛(Monte Carlo)方法在计算风险价值(Value at Risk,VaR)方面的缺陷,文章首先引入GARCH模型来刻画金融数据的波动聚集性,再引入马尔科夫链蒙特卡洛(Markov Chain Monte Carlo,MCMC)方法,来克服GARCH模型参数估计约束条件带来的估计误差。通过对上证50指数的实证分析表明,引入MCMC方法可以提高模型的估计精确度。
