论文部分内容阅读
针对蔬菜总黄酮化学物提取过程复杂、非线性和生物参数难以在线测量等特点,提出了基于改进混合蛙跳算法优化最小二乘支持向量机的蔬菜总黄酮软测量模型。该模型对标准混合蛙跳算法进行改进,采用反向学习的种群初始化策略,确保个体分布的均匀性;并根据群体适应度方差大小,动态调整变异概率,使算法避免陷入局部最优;最后采用改进的混合蛙跳算法对最小二乘支持向量机的参数进行寻优,实现蔬菜总黄酮软测量。仿真结果表明,基于改进混合蛙跳算法的最小二乘支持向量机软测量模型具有测量精度高,稳定性好的优点,有利于蔬菜总黄酮化学物测量工程的实际应用。
In view of the complexity, non-linearity and inaccessibility of biological parameters of on-line measurement of total flavonoids in vegetables, a soft sensorial model for measuring total flavonoids in vegetables was proposed based on improved hybrid frog leaping algorithm and least squares support vector machine. This model improves the standard mixed frog leaping algorithm and adopts the strategy of initial learning of reverse learning to ensure the uniformity of individual distribution. According to the variance of population fitness, the model adjusts the mutation probability dynamically so that the algorithm avoids falling into local optimum. Finally, Improved hybrid frog leaping algorithm optimizes the parameters of least square support vector machine to realize the soft measurement of total flavonoids in vegetables. The simulation results show that the least square support vector machine soft measurement model based on the improved hybrid frog leaping algorithm has the advantages of high measurement accuracy and good stability, which is in favor of the practical application of the total flavonoids chemical measurement project.