基因表达式编程在重金属形态预测建模中的应用研究

ON APPLYING GEP TO MODELLING HEAVY METALS FORM PREDICTION

提出一种改进的GEP(Gene Expression Programming)算法。根据重金属(HM)形态随时间变化(HMFT)的特点,建立基于跳跃基因表达式编程(JM-GEP)的重金属形态预测模型。跳跃算子是该模型的关键。为了保持种群多样性,研究最优保留策略GEP的收敛特性,跳跃算子采用自适应的跳跃概率。针对改进后的JM-GEP算法,提出一种基于GEP的重金属形态预测建模方法。仿真结果表明新模型更适合于HMFT的特性函数,找到全局最优解,且明显优于经典GEP算法及其他算法。该新模型方法还可广泛用于其他时间序列预测问题的研究。

In this paper, we proposed to use an improved GEP （gene expression programming） to build the heavy metal forms prediction model, which is based on jumping genes expression programming （JM-GEP）, according to the characteristic of heavy metal forms changing along with the time （HMFT）. Jumping operators are the key point of the model, in order to keep the diversity of population, we studied the convergence property of optimal retention strategy of GEP. The jumping operators use self-adaptive jumping probability. We put forward a JM- GEP based heavy metals form prediction modelling method aimed at the improved GEP. Simulation results showed that the new model were more fit the characteristic function of HMFT, and found its global optimal solution, and was conspicuously superior to other algorithms. The method of new model could also be widely applied to the researches of other time sequences prediction problems.