中国电机工程学报
中國電機工程學報
중국전궤공정학보
ZHONGGUO DIANJI GONGCHENG XUEBAO
2014年
5期
734-740
,共7页
炉膛火焰监测%火焰图像%特征提取%状态识别%Krawtchouk矩%小波支持向量机%混沌小生境粒子群优化
爐膛火燄鑑測%火燄圖像%特徵提取%狀態識彆%Krawtchouk矩%小波支持嚮量機%混沌小生境粒子群優化
로당화염감측%화염도상%특정제취%상태식별%Krawtchouk구%소파지지향량궤%혼돈소생경입자군우화
furnace flame monitoring%flame image%feature extraction%state recognition%Krawtchouk moment%wavelet support vector machine%chaos niche particle swarm optimization
监测炉膛火焰燃烧状态对防止锅炉爆管起着重要作用。为了进一步提高火焰图像特征提取的准确度和燃烧状态的识别率,文中将Krawtchouk矩引入火焰特征提取,提出了一种将Krawtchouk矩不变量与小波支持向量机相结合的火焰燃烧状态识别方法。首先计算火焰图像的 Krawtchouk矩及Krawtchouk矩不变量,以此构造火焰图像的特征向量;然后根据训练样本的特征向量构造支持向量机,对火焰图像进行状态识别,并采用混沌小生境粒子群算法优化支持向量机中的核函数参数与惩罚因子,使其识别性能最优。大量实验结果表明:与基于 Hu 矩和支持向量机的方法、基于Zernike矩和支持向量机的方法相比,采用Krawtchouk矩不变量作为火焰图像的特征能更好地对火焰图像燃烧状态进行识别,识别率大大提高,且结果与实际情况相符。
鑑測爐膛火燄燃燒狀態對防止鍋爐爆管起著重要作用。為瞭進一步提高火燄圖像特徵提取的準確度和燃燒狀態的識彆率,文中將Krawtchouk矩引入火燄特徵提取,提齣瞭一種將Krawtchouk矩不變量與小波支持嚮量機相結閤的火燄燃燒狀態識彆方法。首先計算火燄圖像的 Krawtchouk矩及Krawtchouk矩不變量,以此構造火燄圖像的特徵嚮量;然後根據訓練樣本的特徵嚮量構造支持嚮量機,對火燄圖像進行狀態識彆,併採用混沌小生境粒子群算法優化支持嚮量機中的覈函數參數與懲罰因子,使其識彆性能最優。大量實驗結果錶明:與基于 Hu 矩和支持嚮量機的方法、基于Zernike矩和支持嚮量機的方法相比,採用Krawtchouk矩不變量作為火燄圖像的特徵能更好地對火燄圖像燃燒狀態進行識彆,識彆率大大提高,且結果與實際情況相符。
감측로당화염연소상태대방지과로폭관기착중요작용。위료진일보제고화염도상특정제취적준학도화연소상태적식별솔,문중장Krawtchouk구인입화염특정제취,제출료일충장Krawtchouk구불변량여소파지지향량궤상결합적화염연소상태식별방법。수선계산화염도상적 Krawtchouk구급Krawtchouk구불변량,이차구조화염도상적특정향량;연후근거훈련양본적특정향량구조지지향량궤,대화염도상진행상태식별,병채용혼돈소생경입자군산법우화지지향량궤중적핵함수삼수여징벌인자,사기식별성능최우。대량실험결과표명:여기우 Hu 구화지지향량궤적방법、기우Zernike구화지지향량궤적방법상비,채용Krawtchouk구불변량작위화염도상적특정능경호지대화염도상연소상태진행식별,식별솔대대제고,차결과여실제정황상부。
Monitoring combustion state of furnace flame plays an important role in preventing boiler tube explosion. To further improve the accuracy of flame image feature extraction and the recognition rate of combustion state, Krawtchouk moment was introduced into feature extraction of flame images, a method of combustion state recognition based on Krawtchouk moment invariants and wavelet support vector machine was proposed. The Krawtchouk moment and Krawtchouk moment invariants of flame images were calculated to construct feature vectors of flame images. Then a support vector machine was constructed according to the feature vectors of training samples to recognize the combustion states of flame. And the kernel parameter and penalty factor of support vector machine were optimized by chaos niche particle swarm optimization algorithm to obtain best recognition performance. A large number of experimental results show that, compared with the method based on Hu moment and support vector machine, the method based on Zernike moments and support vector machine, using Krawtchouk moment invariants as features of flame image can better recognize the combustion state. The recognition rate is greatly improved, and the result is consistent with the actual situation.