【摘要】 利用经交叉验证(CV)方法和遗传算法(GA)优化的支持向量机(SVM)分类模型(CV-GA-SVM)对混合生物质的燃烧特性进行类别预判,并提出综合燃烧特性指数的简便计算式。分类模型以工业分析成分为输入量,以试样标签为输出量,以单生物质数据训练模型。基于单生物质工业分析建立了综合燃烧特性指数的简便计算式。研究表明:CV-GA-SVM模型能够对混合生物质的综合燃烧特性作出准确、快速的预判,正确率为100%,耗用时间为4.4s;简便计算式的计算平均绝对误差为1.68。更多还原

【Abstract】 The support vector machine(SVM)classification model optimized by cross validation(CV)and genetic algorithm(GA)was used to predict the combustion characteristic of mixed biomass,and a simple calculation formula for comprehensive combustion characteristic index was derived,based on the industry analysis of single biomass.The model applies the proximate analysis data as the input and the combustion characteristic based label as the output,and the single biomass data was used to train the model.It is found that,the mixed biomass combustion characteristic can be predicted quickly and accurately by support vector machine model optimized by cross validation and genetic algorithm(CV-GA-SVM),whose accuracy rate is 100% and the running time is 4.4 s;the average absolute error of the formula is 1.68.更多还原