自适应人工蜂群优化极限学习机在拉曼光谱血液定量分析中的应用

Application of Adaptive Artificial Bee Colony Optimization Extreme Learning Machine in Quantitative Analysis of Blood by Raman Spectroscopy

提出了一种基于自适应差分进化人工蜂群优化极限学习机预测血液各组分浓度的方法。首先应用人工蜂群算法对输入权值和隐含层阈值迭代寻优;其次结合差分进化进一步提高模型精度且避免后期易陷入局部最优等问题;由于差分进化算法交叉率和变异率存在凭经验给定的不确定性,最后引入了自适应调整的思想提出自适应差分进化人工蜂群算法优化极限学习机算法的模型,将其应用于血液成分定量分析中。实验表明,自适应差分进化人工蜂群算法优化的极限学习机模型具有较高的预测精度,模型具有较强的稳健性。

A method based on adaptive differential evolution artificial bee colony optimization extreme learning machine is proposed to predict the concentration of each component of blood.First,the artificial bee colony algorithm is used to iteratively optimize the input weights and hidden layer thresholds;secondly,the differential evolution is combined to further improve the model accuracy and avoid problems such as falling into local optimality in the later stage;due to the fact that the crossover rate and mutation rate of the differential evolution algorithm are based on experience,the idea of adaptive adjustment is introduced,and the model of adaptive differential evolution artificial bee colony algorithm to optimize the extreme learning machine algorithm is proposed and applied to the quantitative analysis of blood components.Experiments show that the extreme learning machine model optimized by the adaptive differential evolution artificial bee colony algorithm has high prediction accuracy,and the model has strong robustness.