基于ADE算法优化的木材单板染色全光谱配色模型研究  

Optimization of Full-spectrum Color Matching Model of Wood Veneer Dyeing Based on ADE Algorithm

作　　者：魏艳秀[1] 管雪梅[1] 李文峰 黄青龙 Wei Yanxiu;Guan Xuemei;Li Wenfeng;Huang Qinglong(College of Mechanical and Electrical Engineering,Northeast Forestry University,Harbin Heilongjiang 150040,China)

机构地区：[1]东北林业大学机电工程学院,黑龙江哈尔滨150040

出　　处：《西南林业大学学报（自然科学）》2021年第2期125-132,共8页Journal of Southwest Forestry University：Natural Sciences

基　　金：哈尔滨创新人才研究基金(2013RBFQXJ138)资助。

摘　　要：采用自适应的交叉因子与变异因子在增强全局搜索能力的同时提高收敛速度,通过循环迭代取值的方法确定Stearns-Noechel模型中参数M的最优值。采用新模型对8组标准样进行染色配方预测,以CIEDE 2000色差评价标准对实验结果进行评价。结果表明:基于ADE优化模型预测染色配方得到的拟合样与标准样间ΔE00均小于3;迭代次数是遗传算法的13.04%,差分进化算法的50.00%,寻优速度更快;对于标准样1,新模型预测配方得到拟合样与标准样间ΔE00约为最小二乘法、遗传算法和差分进化算法优化模型的1/5、1/3与1/2,预测精度更高,说明了新模型在染色木材单板全光谱配色方面具有较高的使用价值。The adaptive crossover factor and mutation factor are used to enhance the global search ability and improve the convergence speed.The optimal value of the parameter M in the Stearns-Noechel model is determined by the method of loop iteration.The new model was used to predict the dyeing formula of 8 groups of standard samples,and the experimental results were evaluated by the CIEDE2000 color difference evaluation standard.The experimental results show that the ΔE00 between the fitted sample and the standard sample obtained by predicting the dyeing formula based on the ADE algorithm optimization model is less than 3.The number of iterations is 13.04%of the genetic algorithm and 50.00% of the differential evolution,and the optimization speed is faster.For standard sample 1,the new formula predicts that the ΔE00 between the fitted sample and the standard sample is about 1/5,1/3,and 1/2 of the least squares,GA and DE algorithm optimization models,the prediction accuracy is higher,which shows that the new model has higher value in the full spectrum color matching of dyed wood veneer.

关 键 词：Stearns-Noechel模型 自适应差分进化算法 超定方程组 全光谱配色 木材单板 木材材色 

分 类 号：F302.3[经济管理—产业经济]