论文部分内容阅读
通过从时域和频域提取能够反映火焰稳定性的5个特征量,采用基于实数编码的加速遗传算法的投影寻踪分类模型(RAGA-PPC模型)对这5个特征量进行分析和讨论,得到能够反映高维数据结构的最佳投影方向,给出这5个特征量在评价火焰燃烧稳定性时的重要程度,并根据投影值和样本序号建立火焰燃烧稳定性评价模型,给出评价模型的函数关系式,分析了模型的误差.利用该模型对测试样本进行了预测.结果表明:强度均值这个特征量对火焰燃烧稳定性的影响贡献最小,火焰闪烁频率对火焰燃烧稳定性的影响贡献最大;该评价模型验证了经典的燃烧理论,基本能满足火焰燃烧稳定性判别的要求.
By extracting five feature quantities that can reflect flame stability from the time domain and the frequency domain, the projection pursuit classification model (RAGA-PPC) based on real-coded genetic algorithm is used to analyze and discuss the five feature quantities. Get the best projection direction which can reflect the high dimensional data structure, and give the importance of these five features in the evaluation of the combustion stability of the flame. According to the projection value and the sample number, establish the flame combustion stability evaluation model, and give the evaluation model The error of the model was analyzed and the model was used to predict the test sample.The results showed that the contribution of the mean value of the intensity contribution to the combustion stability of the flame was minimal and the influence of the flame flashing frequency on the combustion stability Maximum; This evaluation model validated the classic theory of combustion, basic to meet the flame combustion stability requirements.