Wide-load operation of large circulating fluidized bed (CFB) units requires coordinated control of in-furnace sulfur capture, selective non-catalytic reduction (SNCR), and wet flue gas desulfurization (WFGD) while balancing emission compliance, reagent consumption, auxiliary electricity use, and operating economy. These coupled stages evolve over different time scales, and their manipulated inputs often change simultaneously under feedback. Fixed-parameter mechanistic models cannot readily accommodate coal-property and equipment-performance drift, whereas prediction-oriented data models may not preserve locally credible input–state–emission responses. This study develops a Physics-Informed Channel-Aware model (PICA) that combines a mechanistic backbone for combustion and pollutant removal with structured data-driven correction along mechanistically supported pathways. The resulting control-oriented hybrid framework improves multistep prediction while retaining credible input–state–emission responses. PICA-EMPC then integrates these predictions with screened local-response and validation-based one-sided uncertainty bounds to define admissible actions, tighten emission constraints, and reallocate removal resources among in-furnace sulfur capture, WFGD, and SNCR. Validation using industrial data from a 350 MW supercritical CFB unit gives coefficients of determination above 0.9 for all pollutant outputs and physical-validity rates of 100% for all manipulated channels. Under identical load, disturbances, and emission constraints, closed-loop validation gives a joint compliance rate of 98.70%, a total operating-cost reduction of 29.55%, and a total input-variation reduction of 26.4%. The results show how physically constrained hybrid modeling can translate prediction into emission-constrained economic decisions for coordinated pollutant-removal operation in a wide-load industrial energy process.

With the increasing integration of renewable energy into the power grid, circulating fluidized bed (CFB) units are undergoing flexibility retrofitting to enhance their rapid load-changing and deep peak regulation (DPR) capabilities, thereby supporting grid stability. However, these transformations present significant challenges for pollutant emission control. To address this issue, a dynamic SO2-NOx emissions model for supercritical CFB units operating under wide load conditions is proposed. This study introduces a comprehensive kinetic mechanism that rigorously represents both homogeneous and heterogeneous reaction pathways associated with combustion and pollutant formation in CFB systems. Integrated prediction models are developed to estimate oxygen concentration, bed temperature, and SO2-NOx emissions across the dense and dilute phases, as well as at key monitoring points. Model validation through closed-loop and step-response testing confirms its predictive accuracy under variable load conditions, effectively capturing the dynamic behavior of pollutant emissions. The results indicate average MAPEs of SO2 and NOx emission concentrations of 14.07 % and 11.64 %, respectively, across three variable load scenarios. Furthermore, an enhanced step generalized predictive control strategy with a feedforward-feedback architecture is implemented. This control framework addresses the high inertia and time delay inherent in emission control systems, enabling optimal real-time regulation of SO2 and NOx emissions. These findings provide actionable insights for achieving high-efficiency, environmentally compliant, and economically viable operation of CFB units.

Circulating fluidized bed (CFB) cogeneration units, as flexible energy sources, face complex interactions of combustion, fluidization, and heat transfer, which limit the full utilization of their energy storage capacity. To enable the integration of renewable generation, it is essential to quantify the energy storage characteristics across a range of operating conditions. In this work, an evaluation method was developed for boiler-side, steam-water, condensate throttling, and heating network energy storage in CFB units. Radiative energy storage coefficients of burning carbon in the lower furnace and of fine circulating bed materials (FCBMs) in the upper furnace were derived from combustion and heat-transfer analyses. The energy storage coefficients of the working fluid and metal surfaces for different steam-water sections were determined based on thermodynamic properties, operational states, and component structures. In addition, a quantitative model for condensate throttling and heating network energy storage was established, and the influence of thermal efficiency variations on the load response rate and available storage time was identified. Through application to a 350 MW supercritical CFB unit, the distribution and dynamics of energy storage across its components were characterized, providing a theoretical foundation for energy storage utilization and rapid load-change control.

To address the pressing need for intelligent and efficient control of circulating fluidized bed (CFB) units, it is crucial to develop a dynamic model for the key operating parameters of supercritical circulating fluidized bed (SCFB) units. Therefore, data-knowledge-driven dynamic model of bed temperature, load, and main steam pressure of the SCFB unit has been proposed. Firstly, A knowledge-driven method is employed to develop a dynamic model for key operating parameters of SCFB units. The model parameters are determined based on the operating data of the unit and continuously optimized in real time. Then, Bidirectional Long Short-Term Memory combined with Convolutional Neural Network and Attention Mechanism is utilized to build the dynamic model of bed temperature, load, and main steam pressure. Finally, a collaboration and integration method based on the critic weight method and the variation coefficient method is proposed to establish data-knowledge-driven model of key operating parameters for SCFB units. The model displays great accuracy and fitting ability compared with other methods and effectively captures the dynamic characteristics, which can provide a research basis for the design of intelligent flexible control mode of SCFB unit.

In order to explore the combustion characteristics of the biomass vibrating grate furnace and realize the control and optimization of the unit combustion process,the mechanism model of the grate combustion process was established through the analysis of the biomass fuel characteristics and combustion mechanism.Moreover,the dynamic change of the grate fuel amount was studied,and the key parameters such as furnace temperature and flue gas oxygen content were predicted.The influence of periodic vibration of grate on combustion state in furnace was discussed.The results show that the amount of fuel in the grate is related to the current feed rate and fuel burning rate.The fuel has a large storage capacity on the grate,which leads to a large delay between the fuel burning and the current feeding.The predicted values of furnace temperature and flue gas oxygen content can follow the measured values well,and their changes are in accord with the combustion characteristics.The periodic vibration of the grate will cause the periodic change of the combustion state in the furnace.When the grate vibrates,the fuel combustion speed,furnace temperature,furnace pressure will increase,and the oxygen content of the flue gas will be reduced.As the vibration of the grate stops,these parameters return to the steady state level.

为响应"十九大"绿色环保精神,满足循环流化床机组超低排放需求,建立准确的NOx排放浓度机理控制模型对于设计循环流化床机组脱硝自动控制方法具有重大意义.从循环流化床锅炉燃烧机理切入,建立即燃碳模型,并将燃料氮分为挥发分氮与即燃碳氮2部分构建NOx炉内自生成模型;考虑CO和即燃碳对NOx的还原作用推导NOx自还原模型;构建选择性非催化还原脱硝模型,综合以上模型建立了适应深度调峰的循环流化床NOx排放模型.探究了机组深度调峰下运行参数与NOx排放浓度的关系以及与选择性非催化还原脱硝效率的影响因素.仿真验证试验表明建立的循环流化床NOx模型取得了较好仿真效果,稳态工况的模型计算值平均预测时间为114 s,与实测值的平均相对误差为2.50%;深度调峰下的模型计算值平均预测时间为126 s,与实测值的平均相对误差为5.42%.模型计算量较实测量提前2～3 min,具有一定预测效果.NOx排放浓度模型可为今后循环流化床机组适应深度调峰、快速变负荷以及超低排放研究提供参考.

当下循环流化床(CFB)机组需参与深度调峰、快速变负荷运行,机组变负荷初期的负荷响应速率主要由其汽水侧蓄热特性决定,因此提出了一种亚临界CFB机组汽水侧蓄热定量计算方法.以某电厂300 MW深度调峰CFB机组为例,根据工质特性对该锅炉汽水流程进行分段,结合锅炉设计数据与实际运行参数,分别计算了不同负荷(30％～100％)工况下各段的工质蓄热系数和金属蓄热系数,并考虑汽轮机热效率变化的影响,分析了 CFB锅炉汽水侧蓄热利用可持续时间与机组负荷响应特性.结果表明:该机组汽水侧蓄热系数随着负荷的降低而增大,在50％负荷以下的变化较大;考虑机组运行稳定裕度差异后,汽水侧蓄热利用可持续时间随着负荷的降低而减小,机组负荷响应能力显著降低.

To achieve the economic and environmentally friendly operation of circulating fluidized bed (CFB) units, it is imperative to conduct optimization to obtain an economical mode of pollutant removal. This article focuses on the multi-objective optimization between SO2-NOx emissions and thermal efficiency for CFB units. According to the operation data and production mechanism of pollutants, models of SO2-NOx emission concentration, bed temperature, and oxygen content based on a convolutional neural network–bidirectional long short-term memory–attention mechanism (CNN-BiLSTM-Attention) were established. Then, an improved quantum genetic algorithm was used to find the optimal input variables of the SO2-NOx emission model. The proposed modeling method was evaluated, and it more accurately simulated the trends of actual operation data than other models under different operating conditions. Combining these models with economic calculations, the operating costs under typical conditions were reduced by 3.20% and 1.82% respectively, and the thermal efficiency increased by 0.72% and 1.07%, which contributes to economical and intelligent operation of the unit.

In order to clarify the dynamic characteristics of bed temperature of biomass circulating fluidized bed（CFB） boiler, so as to establish a CFB combustion control system which is more suitable for biomass, a dynamic bed temperature model is established by analyzing the biomass combustion process and combustion mechanism.On the basis of the theory of instant burning carbon combustion, the correlation degree of temperature field in the furnace is calculated and analyzed. The results show that, the calculated bed temperature can be controlled basically stable near the filtering value of the actual bed temperature, and the variation trend of the bed temperature is similar to that of the actual filter bed temperature, which verifies the adaptability and effectiveness of the model. The temperature correlation difference of the upper and lower parts of the biomass CFB boiler is related to the oxygen content and the temperature of the furnace. The temperature difference of the left and right sides is greatly affected by the flue gas flow. In the upper part of the furnace, the material concentration and the uneven heating surface arrangement are also important reasons affecting the temperature characteristics.

Pollutant prediction for coal-fired circulating fluidized bed units is crucial for ultra-low emission optimization. Accurate prediction models can assist in the control optimization of the unit. Mechanism models are limited by the determination of parameters, coefficients, and fitting functions in the model and require a large amount of operational and unit design data in practical applications. With the development of deep learning, more and more deep learning models are used in parameter prediction. These models suffer from insufficient prediction accuracy when performing parameter prediction tasks due to the lack of a priori knowledge of the mechanism process. This paper analyzed the relationship between the differential equation model under the first-order Taylor expansion and the single-layer Gated Recurrent Unit neural network model. According to the analysis results, this paper proposed a mixed prediction model of SO2 concentration. The ablation study demonstrated the validity of the predictive model structure. The operation datasets of two actual units were used for verification. In terms of MAE indicators, the results of the proposed model on the two data sets are 124.5669 mg/Nm(3) and 178.0473 mg/Nm(3). In terms of MAPE indicators, the results of the proposed model on the two data sets are 5.85% and 14.07%.

In order to analyze the internal combustion mechanism of the biomass circulating fluidized bed boiler and realize the optimization of combustion control, the combustion process and combustion mechanism in the furnace were analyzed through the analysis of the characteristics of biomass fuel and the application of the instantaneous carbon combustion theory, and the combustion process was established. The dynamic change process of carbon burning in the furnace was analyzed to realized the prediction of load, furnace temperature, carbon monoxide emission and flue gas oxygen content. Results show that the combustion rate of instant carbon formed after the fuel devolatilization is slower than that of volatile matter, and the fuel feeding fluctuation directly affects the instant carbon stock in the furnace. As the main source of heat when the feed fluctuates, the oxygen required for instant carbon combustion is less and the combustion is more sufficient. At the same time, the CO valume fraction in the flue gas is lower and the oxygen content is higher. the predicted values of furnace temperature, furnace load, CO valume fraction and flue gas oxygen content are basically the same as the actual values.

Coal slime blending can effectively improve the utilization rate of fossil fuels and reduce environmental pollution. However, the combustion in the furnace is unstable due to the empty pump phenomenon during the coal slurry transport. The combustion instability affects the material distribution in the furnace and harms the unit operation. The bed pressure in the circulating fluidized bed unit reflects the amount of material in the furnace. An accurate bed pressure prediction model can reflect the future material quantity in the furnace, which helps adjust the operation of the unit in a timely fashion. Thus, a deep learning-based prediction method for bed pressure is proposed in this paper. The Pearson correlation coefficient with time correction was used to screen the input variables. The Gaussian convolution kernels were used to implement the extraction of inertial delay characteristics of the data. Based on the computational theory of the temporal attention layer, the model was trained using the segmented approach. Ablation experiments verified the innovations of the proposed method. Compared with other models, the mean absolute error of the proposed model reached 0.0443 kPa, 0.0931 kPa, and 0.0345 kPa for the three data sets, respectively, which are better than those of the other models.

It is urgent to establish a dynamic model of once-through supercritical circulating fluidized bed (SCFB) power generation units in order to explore the flexible operation mode. This study aims to establish such a dynamic model with the balance between model complexity and accuracy. Within the analysis on operation characteristics of SCFB units, some reasonable assumptions and simplifications were intro-duced. Then, a dynamic model structure was derived from mass, energy balance laws as well as ther-modynamic principles. And the unknown parameters in the model were identified by regression analysis and quantum genetic algorithm, combined with running data from a 350 MW SCFB unit. After this, the open-loop step responses and closed-loop experiments were developed to further validate the estab-lished model. The simulation results show that all model outputs can track the dynamic trends of running data in different load conditions with the satisfactory accuracy. More importantly, the model captures the special dynamic characteristics of the SCFB unit. Therefore, the model can be feasible and applicable for advanced control algorithms test and flexible operation control system design of SCFB units. (C) 2021 Elsevier Ltd. All rights reserved.

As the emission regulation becomes more stringent, in-situ desulfurization and wet flue gas desulfurization have been widely used for circulating fluidized bed (CFB) boilers. However, few studies focus on real-time operation optimization of combined desulfurization system (CDS) in wide-load range based on in-situ SO2 emission prediction. In this work, according to mechanism analysis, a dynamic prediction model of in-situ SO2 emission was developed. Then, closed-loop and step response experiments were conducted to validate the established model, combined with measured data from a 300 MW CFB boiler. Validation results show that the model has satisfactory prediction accuracy in different load conditions and it can capture special dynamic characteristics of in-situ SO2 desulfurization. More importantly, the model can be feasible for building an economical operation strategy of CDS. Thus, a novel control strategy based on dynamic prediction model was designed, which contributes to decrease the SO2 emission fluctuation and operation costs by correcting the limestone-feed rate in advance. The proposed control strategy was applied to actual operation of the 300 MW CFB boiler. In the long-term operation of CDS after optimization, it was easier to reach ultra-low SO2 emission standard and limestone consumption reduces 15.14% compared with before the operation optimization.

为提高循环流化床机组的变负荷速率,基于火电机组锅炉侧和汽轮机侧响应存在差异,提出了基于凝结水节流和热网蓄能利用的快速变负荷方式.首先建立了抽汽系统以及汽轮机的机理模型,并计算了多个工况下负荷的理论值,验证上述模型的准确性;然后对抽汽系统蓄热进行分析,建立了基于凝结水节流的除氧器蓄热和热网蓄热定量计算模型,分析了影响循环流化床供热机组蓄热能力主要因素;最后提出了蓄热和负荷等响应特性.结果表明:循环流化床供热机组因存在冷渣环节,可调节凝结水流量受限,除氧器蓄热较煤粉炉更小;热网蓄热作为一个巨大的蓄热体,充分利用可提供可观的负荷提升,且短时间内的波动对于热网整体并无影响;除氧器蓄热和热网蓄热2种方式结合使用,在带供热工况下,变负荷时间可以缩短达4 min左右.

As the proportion of renewable energy generation increases in power grid, most of circulating fluidized bed (CFB) boiler-turbine units are required to operate in a wide-load range to maintain the stability of power grid. However, few studies focus on modeling of CFB units operated in a wide-load range, especially in ultra-low load. This study develops a dynamic model for subcritical CFB boiler-turbine units operated in a wide-load range, which can provide foundation for designing flexible operation mode. In this work, different characteristics of subcritical CFB units in ultra-low load were analyzed. Then, the model structure with balance between complexity and accuracy was derives from mass and energy conservation laws as well as thermodynamic principles. The parameters and functions in model were identified based on running data from a 330 MW subcritical CFB unit. Validation results show that the of model has satisfactory accuracy in wide-load range, from ultra-low load to full load. Besides, the model can reflect essential dynamic characteristics of the unit. Therefore, the model can be feasible for performance analysis and testing control strategy.

Large-scale circulating fluidized bed (CFB) power generation units with high steam pressure and temperature have been put into operation in order to obtain high cycle efficiency and less emissions. To ensure the flexible operation of such CFB units, it is necessary to build a model for the design of coordinated control system (CCS). However, few studies focus on modeling of such CFB units. Thus, this work develops a mechanistic dynamic nonlinear model of a large-scale supercritical CFB unit, which is suitable for CCS design because of its relatively low complexity. Firstly, model structure is derived from mass and energy conservation laws, combined with analysis of operational characteristics of a 600 MW supercritical CFB unit. Then, unknown parameters and nonlinear functions in model are identified based on running data by using regression analysis and optimization algorithm. Validation results show that the model has satisfactory accuracy and it can capture essential dynamic characteristics of the unit and nonlinearity of CCS. More importantly, the model accuracy is further improved by proposed variable dynamic parameters and correction of coal quality. After this, transfer function matrix of CCS is derived from dynamic nonlinear model, which can be applied to controller design and simulation analysis.

With the stricter emission regulation taking effect, most of circulating fluidized bed (CFB) power generation units operates with conventional desulfurization in CFB furnace and flue gas desulfurization (FGD) technology to meet the requirements of ultralow SO 2 emission. Therefore, it is urgent to analyse operating cost of the combined desulfurization system and explore the low-cost operating mode for CFB units. In this paper, the operating cost model of combined desulfurization system was established, in which the influence of desulfurization in furnace on thermal efficiency of the CFB boiler was considered. Then, the operating cost of combined desulfurization system under different load conditions for a 300MW subcritical CFB unit was analysed, combined with measured data. The operating cost optimization result shows that, under different loads, when Ca/S is in the range of 0.4 to 0.5, the operating cost of combined desulfurization system reaches the minimum. In addition, as the unit load increases, the higher proportion of SO 2 removal in FGD is required to achieve low-cost operating mode for CFB units.

Under such a circumstance that the scale of renewable power connected into grids increases companied with more fluctuation, the flexibility and stability in power generation have been focus. Circulating fluidized bed (CFB) has unique merits in deep peak shaving, but its operation presents multi-influencing factors and multi-mode characteristics, which makes it very difficult to monitor the operation state. Toward this end, a novel performance evaluation framework has been proposed. The proposed framework contains two main parts: deep feature extraction conducted by deep belief networks (DBN), connecting with performance status classification by least square support vector machine (LSSVM). In this framework, massive operation data detected by sensors and reference status labels were entered into DBN for dimension reduction and feature extraction in a semi-supervised way. LSSVM finished the status classification based on these features. The final classification results are processed by DBN and LSSVM successively, which can not only make full use of the multidimensional parameters of CFB, but also avoid the influence of multimode of CFB. Besides, some comparations of the case study are conducted and analysed respectively to verify the efficiency and accuracy of the performance evaluation framework.

To meet the stringent emission standards and achieve cleaner production of circulating fluidized bed units, it is necessary to build a dynamic model of pollutants emission for creating an economical and environmentally friendly pollutant removal operation mode. This article fources on the modeling and accurate prediction of SO2-NOx emission concentration of the circulating fluidized bed unit. According to the generation and reduction mechanism of pollutants, the model inputs are selected and determined by Pearson coefficient. Then, a dynamic model of SO2-NOx emission concentration based on extreme learning machine is developed, and quantum genetic algorithm is used to optimize the connection weight between the input layer and the hidden layer and threshold of the extreme learning machine, which contributes to increase prediction performance. The test result shows that the optimized model can effectively imitate the dynamic trends of actual measured data with appropriate accuracy, the mean absolute percentage error of SO2 concentration and NOx concentration are 4.63% and 3.09% respectively. And the model's satisfactory generalization ability is demonstrated based on the generalization experiment. In addition, compared with other methods, the proposed modeling method for SO2-NOx concentration has more suitability and accuracy under dynamic conditions, which contributes to online optimization of pollutant control and intelligent development of circulating fluidized bed units.