基于支持向量机和遗传算法的燃煤电站锅炉多目标燃烧优化

本文采用支持向量机方法建立了350MW燃煤电站锅炉NOx预测模型和锅炉效率预测模型,并采用遗传算法对NOx和锅炉效率进行多目标优化,表明支持向量机和遗传算法可以用于指导参数调节,进行燃烧优化。     

基于支持向量机和遗传算法的燃煤电站锅炉多目标燃烧优化

采用支持向量机方法建立350MW燃煤电站锅炉NOx预测模型和锅炉效率预测模型,并采用遗传算法对NOx和锅炉效率进行多目标优化,表明支持向量机和遗传算法可以用于指导参数调节,进行燃烧优化。     

燃煤锅炉NO_x排放浓度的智能预报

在热态试验数据的基础上,分别应用BP(神经网络)和SVM(支持向量机)回归算法建立了燃煤机锅炉NOx排放特性模型,并验证了模型的准确性。结果表明,BP网络模型对检验样本的最大预测误差、最小预测误差和均方差分别为4.263%、0.556%和2.2133%,支持向量机模型对检验样本的最大预测误差、最小预测误差和均方差分别为2.121%、0.091%和0.4549%。两种智能技术都能对锅炉在不同工况下的NOx排放做出较为准确的预报,但支持向量机在泛化能力、收敛速度、最优性等方面明显优于神经网络。     

在线监测燃煤锅炉NOx排放的自适应支持向量机模型

提出了一种基于改进的在线支持向量机自适应建模方法,并应用于电站锅炉NOx排放连续监测和特性分析.对常规在线支持向量机方法进行了改进,提出了新的样本剔除规则,保证了训练集内样本分布的均匀性.通过该改进方法对基于试验数据的常规向量机模型预测余差进行了连续估计,并预估煤质等因素引起的NOx排放特性的变化,从而补偿了实际工况与试验工况的差别,以便正确给出锅炉NOx排放特性.     

基于支持向量机的燃烧过程飞灰含碳量与NOx排放特性模型

大型四角切圆电站锅炉NOx排放和飞灰含碳量是燃烧优化的两个方面,也是电厂关心的重要问题。影响两者的因素众多而且复杂,对锅炉飞灰含碳量和NOx排放特性进行建模是实现燃烧优化的一个前提。文中首先利用交叉验证算法分析样本数据,对支持向量机中的参数C和σ进行选择,再应用支持向量机理论建立了飞灰含碳量和NOx排放特性模型,最后利用试验数据对模型进行了校验。研究结果表明,采用支持向量机算法建模误差较小,达到了比较准确的预测效果。     

电站锅炉NOx排放与效率的混合建模及优化

由于锅炉设备庞大,运行条件复杂,煤种多变等因素,很难建立锅炉NOx排放与效率的函数模型。利用最小二乘支持向量机(LS-SVM)建立了以锅炉NOx排放与热效率为输出的混合模型,并对此模型进行了校验。结果表明,该模型具有调节参数少、运算速度快、结果稳定、预测精度高等优点,可以根据燃煤特性以及各操作参数准确预报锅炉在不同工况下的NOx排放和效率。针对模型的多目标优化问题,采用多目标粒子群优化算法MOPSO(multiple objective particle swarm optimization)对某工况进行优化仿真,在提高效率的同时降低了NOx排放。     

基于共享最小二乘支持向量机模型的电站锅炉燃烧系统的优化

电站锅炉燃烧系统是一个复杂的多输入多输出系统,为了在同一个模型中实现高效率、低污染物排放的优化目标,对标准最小二乘支持向量机回归方法进行了扩展．借助某电厂1000MW超超临界锅炉的现场燃烧调整试验数据,建立了以锅炉热效率和NOx排放质量浓度为输出的共享最小二乘支持向量机（LSSVM）模型,采用一种改进的粒子群算法对共享模型中的锅炉运行工况进行了寻优．结果表明：在共享LSSVM模型中,锅炉热效率和NOx排放质量浓度的平均预测误差分别可达到0．028％和2．16％,搜索得到的高效率和低NOx排放的参数组合可为电站锅炉优化运行提供指导．     

四角切圆燃煤锅炉低NO_x燃烧优化研究

借助锅炉燃烧特性试验数据,建立了基于支持向量回归的四角切圆燃煤锅炉NOx排放模型.经过训练和校验,并与神经网络模型进行对比,结果表明:SVR模型更加适合于实炉热态测试工况较少的小样本学习,而且其精度能够满足工程的实际要求,能够较为准确的对不同工况下的电站锅炉NOx排放特性进行预测.在获得该模型的基础上,结合全局寻优的遗传算法,以锅炉的运行调节参数为优化目标函数的自变量,对NOx排放进行寻优,并获得了具体工况下的最佳操作参数.     

基于万有引力搜索算法的电厂锅炉NO_x排放模型的参数优化

以某330MW煤粉汽包锅炉为测试对象,以支持向量回归机为基础,利用最小二乘支持向量机和一种新的寻优算法———万有引力搜索算法进行了综合建模及参数优化.结果表明:建立的模型较好地实现了对电厂锅炉NOx质量浓度排放量的预测;与遗传算法、蜂群算法和粒子群算法相比,万有引力搜索算法能更好地找到未知优化参数,使得所建模型具有更高的预测能力和泛化能力,从而有效控制燃煤电厂NOx的排放量.     

基于SVM的燃煤电站锅炉飞灰含碳量预测

将支持向量机方法引入燃煤电站锅炉飞灰含碳量预测领域．该预测方法很好地建立了燃煤电站锅炉飞灰含碳量特性与运行参数之间的复杂关系模型,并考虑到运行参数之间的耦合性,具有预测能力强、全局最优及泛化性好等优点．将该方法应用于某300 MW燃煤电站锅炉中,经过训练后的SVM模型对检验样本飞灰含碳量进行预测,均方根误差和平均相对误差分别为1．39%和1．30%,相当于BP网络模型的22．20%和21．07%．应用结果表明,支持向量机方法优于多层BP神经网络法,能很好地满足预测要求．     

催化燃烧对均质压燃发动机排放影响的数值模拟

通过耦合DETCHEM软件包及CHEMKIN软件包中的SENKIN模块,对活塞顶涂有催化剂的均质压燃(HCCI)发动机的燃烧过程进行了数值计算,建立了多区模型.利用此模型分析了催化燃烧对HCCI发动机缸内温度、热释放速率以及未燃碳氢化合物(UHC)、氮氧化合物(NOx)、一氧化碳(CO)排放的影响,结果表明催化燃烧能降低UHC、CO的排放,但NOx的排放会有所升高.对不同催化剂及混合催化剂对HCCI发动机缸内温度、热释放速率以及UHC、NOx、CO排放的影响进行了探索,结果表明,和金属铂相比,以铑作催化剂时UHC的排放降低,但NOx、CO排放会有所升高;采用500/0Pt-500/0Rh的混合催化剂时,UHC、NOx的排放介于1000/0Pt与1000/0Rh之间,但CO的排放却比采用1000/0Pt与1000/0Rh时都要低.     

旋流对天然气高温空气燃烧特性影响的数值模拟

以工业炉的高温空气燃烧技术应用为背景,对一个新型轴向旋流式单烧嘴燃烧室内天然气的高温空气燃烧特性进行了数值研究。采用数值模拟的方法研究了同心式轴向旋流燃烧器（HCASbumer）中螺旋肋片的旋转角度对燃烧特性的影响,其中湍流采用Reynolds应力模型,气相燃烧模拟采用β函数形式的PDF燃烧模型,采用离散坐标法模拟辐射换热过程,NOx模型为热力型与快速型。计算结果表明,对预热空气采用旋转射流时,能明显降低NOx生成量。对于HCAS型燃烧器,随着空气射流旋转角度的增大,燃烧室内的回流区域增大增强,降低了局部的氧体积分数分布,燃烧室中平均温度和最高温度都有所增加,且燃烬程度大幅度提高,而局部高温区缩小,只在靠近入口处出现。总的NOx排放量随着空气射流旋转角度的增大先减小,后增大。因此,适当调整肋片的旋转角度可以降低NOx生成量。     

空气预热温度对燃烧状况的影响

空气预热是有效的节能技术,但预热温度的提高同时带来NOx排放浓度增加的问题。为了了解其规律,本文针对某烯烃厂芳烃加热炉的空气预热改造项目,对不同空气预热温度情况下的燃烧状况和NOx排放规律做了研究。首先利用数值模拟方法,构建了加热炉三维几何模型,将燃烧模型和NOx生成模型结合,对不同空气温度下的燃烧温度和NOx排放进行模拟,对炉膛内部温度分布及NOx排放规律做了研究,最后找出空气预热最佳温度。     

运行调整对循环流化床锅炉NOx排放影响分析

在300 MW循环流化床锅炉上进行了运行调整对NOx排放影响分析,考查了燃烧温度、过剩空气系数、空气分级、Ca/S等因素对NOx排放的影响.试验结果表明：燃烧温度与过剩空气系数对NOx的排放浓度影响显著;空气分级燃烧对NOx的排放浓度有一定影响,但影响相对有限;过大的Ca/S易导致NOx排放浓度的增加.     

Comparison of staged combustion properties between bituminous coals and a low-rank coal; Fiber-shaped crystallized carbon formation,NOx emission and coal burnout properties at very high temperature

Previously, we developed a new-concept for a drop-tube furnace in order to investigate staged combustion properties for pulverized coals. Two high-temperature electric furnaces were connected in series. Coal was burnt under fuel-rich conditions in the first furnace, then, staged air was supplied at the connection between the two furnaces. In the present study, we investigated influence of burning temperature on NOx emission and combustion efficiency by using the furnace. The influence of the temperature differed between hv-bituminous coals and a sub-bituminous coal. For the hv-bituminous coals, combustion efficiency was improved when burning temperature in the fuel-rich region rose. When combustion efficiency was improved, NOx emission decreased. The NOx reduction reaction in the fuel-rich region was promoted by increasing the burning temperature in this region. On the other hand, NOx emission increased for the sub-bituminous coal when the temperature was higher than 1800 K. Usually, combustion efficiency was increased with burning temperature. However, combustion efficiency lowered for the sub-bituminous coal when burning temperature was higher than 1900 K. We observed the ash obtained by this temperature condition using Scanning Electron Microscope (SEM) and, Transmission Electron Microscope (TEM) and observed fiber shaped carbon. The difference in NOx properties was derived as a difference of hydrocarbon concentration. For low-rank coals (sub-bituminous or lignite), the hydrocarbon formation rate was smaller than that for hv-bituminous coals. When the hydrocarbon contribution to the NOx reduction reaction was large, NOx emission decreased with increasing burning temperature; however, hydrocarbon content in volatile matter was small for low-rank coals.     

Unreacted-core model applied on the NOx emission of coke combustion

Despite numerous experimental studies of the NOx emission of coke combustion, there is no good computational model for NOx emission of the coke combustion. In this study, experimental studies of exhaust emissions of coke combustion were conducted by using micro-sinter equipment. Unreacted-core model was applied to kinetic analysis. The calculated NOx content curve described the measured data well, and the peak and integral total NOx difference values between the calculated and experimental data were 2.7 ppm and 0.15 mL, respectively, which illustrated successful application of unreacted-core model for the NOx emission of coke combustion. The equation also illustrated that particle size of coke breeze and O₂ concentration had a great influence on the NOx emission of coke combustion.     

天然气燃料轴向分级预混燃烧特性研究

低NOx排放是燃气轮机燃烧室的重要性能指标,面对燃烧室出口温度不断增加的趋势,新型燃烧技术探索应用成为必然。燃料轴向分级(Axial Fuel Staging,AFS)燃烧作为一项可行技术方案已在各燃机厂商的最先进燃气轮机燃烧室上获得应用,其主要通过降低高温烟气有效停留时间和低氧浓度燃烧来实现低NOx排放。基于Chemkin平台建立轴向分级预混燃烧室化学网络模型,针对1 973 K燃烧室,研究二级负荷比例、当量比和停留时间对NOx/CO排放的影响规律,对比分析主燃区高温烟气与二级未燃预混气掺混特征的影响,获得AFS燃烧的污染物排放特性和关键影响因素。同时,在贫预混燃烧器上,设计二级喷射段,实验研究二级火焰结构、污染物排放等燃烧特性。结果表明,相比于常规贫预混燃烧,AFS燃烧在高温区体现出很好的低NOx优势,且能拓宽低NOx工况范围,其中主燃区温度、二级当量比和停留时间匹配特征、主燃区高温烟气与二级预混气掺混性能等是关键。     

Mechanism analysis on the pulverized coal combustion flame stability and NOx emission in a swirl burner with deep air staging

Low NOx burner and air staged combustion are widely applied to control NOx emission in coal-fired power plants. The gas-solid two-phase flow, pulverized coal combustion and NOx emission characteristics of a single low NOx swirl burner in an existing coal-fired boiler was numerically simulated to analyze the mechanisms of flame stability and in-flame NOx reduction. And the detailed NOx formation and reduction model under fuel rich conditions was employed to optimize NOx emissions for the low NOx burner with air staged combustion of different burner stoichiometric ratios. The results show that the specially-designed swirl burner structures including the pulverized coal concentrator, flame stabilizing ring and baffle plate create an ignition region of high gas temperature, proper oxygen concentration and high pulverized coal concentration near the annular recirculation zone at the burner outlet for flame stability. At the same time, the annular recirculation zone is generated between the primary and secondary air jets to promote the rapid ignition and combustion of pulverized coal particles to consume oxygen, and then a reducing region is formed as fuel-rich environment to contribute to in-flame NO_X reduction. Moreover, the NOx concentration at the outlet of the combustion chamber is greatly reduced when the deep air staged combustion with the burner stoichiometric ratio of 0.75 is adopted, and the CO concentration at the outlet of the combustion chamber can be maintained simultaneously at a low level through the over-fired air injection of high velocity to enhance the mixing of the fresh air with the flue gas, which can provide the optimal solution for lower NOx emission in the existing coal-fired boilers.     

A comprehensive study on NOx emission and fuel nitrogen conversion of solid biomass in bubbling fluidized beds under staged combustion

Despite the relatively low emissions in fluidized-bed combustion, NOx emission for biomass combustion is still a major concern because of increasingly stricter regulations. To realize NOx emission behavior in fluidized beds comprehensively, the effects of bed temperature, excess oxygen, staged combustion, and flue gas recirculation (FGR) are investigated in this study. In particular, three different types of operation are applied in staged combustion to find out the key parameter. The results indicate that NOx emissions increase with both bed temperature and excess oxygen, in which the influence of excess oxygen is greater than the other. Lowering bed temperature by water addition seems to be able to simultaneously reduce NOx emission and agglomerate formation, especially for fuels with high nitrogen content, but the pros and cons should be considered. The results in staged combustion infer that the residence time is much more critical than the stoichiometry in the bed. As for FGR, its impact appears to depend on the type of fuel. The correlation between NOx emission behavior and fuel characteristics is also scrutinized; it is concluded that the fuel-N conversion to NOx is essentially related to some features of fuels.     

CFD study on influence of fuel temperature on NOx emission in a HiTAC furnace

The influence of the fuel temperature on NOx formation was investigated numerically. For this purpose CFD modeling of NOx emission in an experimental furnace equipped with high temperature air combustion (HiTAC) system was studied. The comparison between the predicted results and measured values have shown good agreement, which implies that the adopted combustion and NOx formation models are suitable for predicting the characteristics of the flow, combustion, heat transfer, and NOx emissions in the HiTAC chamber. Moreover the predicted results show that increase of the fuel temperature results in a higher fluid velocity, better fuel jet mixing with the combustion air, smaller flame and lower NOx emission.