An Extended Self-Organizing Map Network for Modeling and Control of Pulse Jet Fabric Filters

ABSTRACT

Pulse jet fabric filters (PJFFs) have become an attractive option of particulate collection utilities, because they can meet stringent particulate emission limits regardless of variation in operating conditions. Despite their wide applications, the present control algorithm for PJFFs can best be described as rudimentary. In this paper, a modeling and control strategy based on the local model network (LMN) is proposed. An extended self-organizing map (ESOM) network is developed to construct the LMN model of the filtration process using the filter's input-output data. Subsequently, these ESOM local models are incorporated into the design of local generalized predictive controllers (GPC), and the proposed controller design is obtained as the weighted sum of these local controllers. Simulation results show that the proposed controller design yields a better performance than both conventional GPC and proportional plus integral (PI) controllers yield.     

The Potential of Pulse-Jet Baghouses for Utility Boilers. Part 2: Performance of Pulse-Jet Fabric Filter Pilot Plants

Pulse-jet fabric filters (PJFFs) are widely used in U.S. industrial boiler applications and in utility and industrial boilers abroad. Their small size and reduced cost relative to more conventional reverse-gas baghouses makes the use of PJFFs appear to be an attractive particulate control option for utility boilers. This paper (Part 2 of a three-part series) summarizes the results of pilot PJFF studies sponsored by the Electric Power Research Institute at different utility sites in the United States. The purpose of these tests is to evaluate PJFF performance for U.S. fossil-fuel-fired applications. These data are also used to corroborate the results of a recent worldwide survey of PJFF user experience, as described in Part 1 of this series. Part 3 will provide a cost comparison of PJFFs to other particulate control options such as electrostatic precipitators and reverse-gas baghouses.     

The Potential of Pulse-Jet Baghouses for Utility Boilers Part 3: Comparative Economics of Pulse-Jet Baghouse,Precipitators and Reverse-Gas Baghouses

Pulse-jet fabric filters (PJFFs) are widely used in U.S. industrial applications, and in both utility and industrial boilers abroad. Their smaller size and reduced cost relative to more conventional baghouses make PJFFs an attractive particulate control option for utility boilers. This article which is the third in a three-part series, compares the cost of PJFFs with electrostatic precipitators (ESPs) and reverse-gas baghouses (RGBs).

This article presents the capital, operating and maintenance (O&M), and level-ized costs for ESPs, RGBs and PJFFs. The particulate control equipment design and pricing are supplied by manufacturers of the control equipment. A comparison of costs for a base case 250-MW boiler indicates that the PJFF capital cost is 22 percent lower than the cost of an ESP with 400 SCA and 12-inch plate spacing; in addition the PJFF is 35 percent lower than the cost of an RGB. The levelized cost for a PJFF is about equal to the cost of the ESP but 14 percent lower than the cost of the RGB. Overall, the attractiveness of a PJFF versus an ESP depends on the coal type and the outlet emissions limit required. PJFF is favored when low-sulfur coal is fired due to the high-resistivity fly ash. Also, PJFF is favored as more stringent outlet emission rates are required.     

A quantitative approach to the traffic air quality problem: the Traffic Air Quality index

The Traffic Air Quality (TAQ) model is a simple tool to estimate traffic fine particulate emissions on roadways (g/km) and can be used for both real-time analysis and for localized conformity analysis ("hot-spot" analysis for nonattainment areas) as defined by 40 CFR 93.123. This paper is a follow-up to a study published earlier regarding the development of the TAQ model. This paper shows how local air quality levels can be a factor in traffic management in nonattainment areas. Similar to the industrial source quotas measured in tons per year, it is proposed that road segments are to be assigned emission quotas (or TAQ indices) measured in pollutant mass emitted per road length (g/km) above which traffic-measures have to be taken to reduce the fine-particulates emissions on such road links. The TAQ model as well as traffic-rerouting measures along with the Intelligent Transportation System (ITS) protocols can be used to have a real-time control of the traffic conditions along expressways to maintain the fine-particulates emissions below the quota assigned per road link and consequently improving the over all local air quality in nonattainment areas.     

Design of optimized PI controller for 7-level inverter: a new control strategy

Photovoltaic (PV) system produces electricity that differs from variations in environmental parameters such as temperature and solar radiation. The PV network will operate at maximum power point (MPP) and deal with an ever-increasing energy demand, that changes from both load and weather conditions.” Moreover, energy storage devices could be a potential solution for improving the efficiency and performance of renewable energy sources (RES). This paper intends to establish a control design by an optimization-assisted PI controller for a 7-level inverter. Accordingly, the gains of PI controller are adjusted dynamically by FireFly Integrated-Sea Lion Optimization algorithm (FFI-SLnO) that integrates the concepts of both Sea Lion Optimization (SLnO) and FireFly algorithm (FF). The gains should be tuned such that the error among the reference signal and fault signal should be low and hence better dynamic performance can be obtained by the presented optimized PI controller. Finally, the performance of the proposed method is compared over other traditional models with respect to certain measures and its superiority is proved.

     

模糊神经网络控制在污水处理中的应用研究

对模糊神经网络（FNN）技术应用于污水处理进行了研究，提出了一个具有五层的模糊神经网络控制器，仿真实验表明，该控制系统具有很强的鲁棒性与容错性。该控制器能够自动调整隶属度函数、动态优化控制规则，将其应用于溶解氧控制和出水水质预测，结果表明可以快速、有效地使溶解氧浓度达到期望值，实际出水水质预测结果也具有很好的收敛性和预测精度。     

模糊神经网络控制在污水处理中的应用研究

对模糊神经网络(FNN)技术应用于污水处理进行了研究,提出了一个具有五层的模糊神经网络控制器,仿真实验表明,该控制系统具有很强的鲁棒性与容错性.该控制器能够自动调整隶属度函数、动态优化控制规则,将其应用于溶解氧控制和出水水质预测,结果表明可以快速、有效地使溶解氧浓度达到期望值,实际出水水质预测结果也具有很好的收敛性和预测精度.     

A Quantitative Approach to the Traffic Air Quality Problem: The Traffic Air Quality Index

Abstract

The Traffic Air Quality (TAQ) model is a simple tool to estimate traffic fine particulate emissions on roadways (g/km) and can be used for both real-time analysis and for localized conformity analysis (“hot-spot” analysis for nonattainment areas) as defined by 40 CFR 93.123. This paper is a follow-up to a study published earlier regarding the development of the TAQ model. This paper shows how local air quality levels can be a factor in traffic management in nonattainment areas. Similar to the industrial source quotas measured in tons per year, it is proposed that road segments are to be assigned emission quotas (or TAQ indices) measured in pollutant mass emitted per road length (g/km) above which traffic-measures have to be taken to reduce the fine-particulates emissions on such road links. The TAQ model as well as traffic-rerouting measures along with the Intelligent Transportation System (ITS) protocols can be used to have a real-time control of the traffic conditions along expressways to maintain the fine-particulates emissions below the quota assigned per road link and consequently improving the over all local air quality in nonattainment areas.     

ANFIS在混凝投药前馈控制器中应用的仿真研究

在水处理中混凝投药前馈控制器的应用效果好坏关键在于控制器是否对混凝投药过程具有良好的模型辨识能力,传统的控制器效果都不太理想,而且存在沉淀池出水浊度波动大,药剂浪费严重等问题。为了解决该问题,介绍了一种用多层前馈神经网络优化模糊逻辑系统的自适应模糊推理系统——ANFIS。它具有良好的非线性函数逼近能力,在ANFIS投药前馈控制器的设计中,运用减法聚类对样本数据进行空间划分,获取初始模糊隶属函数和模糊规则,得到ANFIS模型的初始结构。用烧杯试验历史数据进行了仿真验证,并与传统的回归模型前馈投药控制仿真比较,结果表明ANFIS投药前馈控制模型明显优于回归模型,它能够根据原水水质适时有效预测混凝投药量。     

The Impact of Particulate Emissions Control On the Control of Other MWC Air Emissions

On December 20, 1989, the Environmental Protection Agency (EPA) proposed revised new source performance standards for new municipal waste combustion (MWC) units and guidelines for existing sources. The proposed national regulations require tighter particulate matter control and address pre-combustion, combustion, and post-combustion controls, the latter two depending on capacity and age of the facility.

The air pollutants of concern when municipal solid waste (MSW) is burned will be discussed. Generally, particulate control is an inherent part of the systems used to limit the emissions of these air pollutants. The relationships between MWC air emissions (acid gases, trace organics, and trace heavy metals) control and particulate control will be discussed. Test results to quantify air pollutant emissions from MWC units and their control will be presented and compared with the proposed regulations.     

Consultant Guide/1989

The European Economic Community (EEC) has proposed strict limits on emissions of dioxins and furans from hazardous waste incinerators. The proposed limit is 0.1 ng/Nm³, expressed as the 2,3,7,8 TCDD toxic equivalent of 17 specific dioxin and furan congeners. These limits will potentially redefine technology selection and design for combustion, energy recovery, and air pollution control. The EPA has a different approach for controlling emissions of products of incomplete combustion (PICs) and reformation products such as dioxins and furans. Rather than limiting these contaminants individually and quantitatively, EPA proposes controlling them by assuring good combustion as measured by stack emissions of carbon monoxide (CO) and total hydrocarbons (THC).

Dioxins and furans are combustion by-products and emission control relies mainly on control of the combustion process. These compounds can also be reformed from certain precursor compounds and elements in lower temperature regions of the system downstream of the combustion process. Air pollution control technologies have demonstrated the ability to remove dioxins and furans as contaminants on fine particulate.

This paper will discuss the two regulatory approaches, the mechanisms for the formation and reformation of dioxins and furans, and the technologies available to control emissions.     

1985 Update,Operating History and Current Status of Fabric Filters in the Utility Industry

Stringent particulate emissions limits and increasing awareness of stack opacity is leading the utility industry to use high efficiency particulate control systems. In response to this trend, the Electric Power Research Institute (EPRI) is conducting several research programs aimed at improving the operation, maintenance and performance of particulate control systems. One of these programs, RP-1401, “Reliability Assessment of Particulate Control Systems,” is developing operation, maintenance and design data bases for both electrostatic precipitators and fabric filters. This paper discusses some of the intermediate findings of the work done on fabric filters.     

The Atmospheric Dust Problem

This article concerns the control of emissions from a local dolomite plant through a series of corrective measures. With a new State regulation on particulate emissions (process weight table) passed in 1965, and with a new local control program, a goal was established toward which control measures were directed. Inplant investigations revealed that measures would be necessary to eliminate emissions from the pulverizer and conveyor belts. In order to meet State requirements, a better method of collection for the dryer was required. In order to evaluate the effect of these changes on the ambient air, dustfall jars and high-volume air samplers were used and stack sampling was conducted at and around the local dolomite plant and also at another such plant in Venice, Fla. Through successive control and testing, emission factors for dolomite plants, as well as reduction of plant emissions and resident complaints, were established.     

Status of the DOE/GEESI In-Duct Scrubbing Pilot Study

The recent promulgation of stack height regulations and possible changes in the National Ambient Air Quality Standards for sulfur dioxide and the associated dispersion model methodologies, could require older power plants to develop new compliance strategies and upgrade emission control systems. In such situations, an inexpensive, moderate removal efficiency flue gas desulfurization technology could maintain the cost effectiveness of these plants. Such a technology was selected by the Department of Energy for demonstration in its Acid Rain Precursor Control Technology Initiative. The process applies the rotary atomizer techniques developed for lime slurry dry flue gas desulfurization spray absorbers, and utilizes existing ductwork and particulate collectors. This induct scrubbing technology is anticipated to result in a dry desulfurization process of moderate removal efficiency. The critical elements for successful application are (i) adequate mixing for efficient reactant contact, (ii) sufficient residence time to produce a non-wetting product, and (iii) appropriate ductwork cross sectional areas to prevent deposition of reaction products before drying. The ductwork in many older power plants, previously modified to meet 1970 Clean Air Act requirements for particulate control, usually meets these criteria. A pilot study of the process is now in design-construct phase. Testing will start in 1987 and determine conditions under which the technology would be a cost effective approach to meeting emission reduction design criteria     

Neural network and multiple regression models for PM10 prediction in Athens: a comparative assessment

Particulate atmospheric pollution in urban areas is considered to have significant impact on human health. Therefore, the ability to make accurate predictions of particulate ambient concentrations is important to improve public awareness and air quality management. This study examines the possibility of using neural network methods as tools for daily average particulate matter with aerodynamic diameter <10 microm (PM10) concentration forecasting, providing an alternative to statistical models widely used up to this day. Based on a data inventory, in a fixed central site in Athens, Greece, ranging over a two-year period, and using mainly meteorological variables as inputs, neural network models and multiple linear regression models were developed and evaluated. Comparison statistics used indicate that the neural network approach has an edge over regression models, expressed both in terms of prediction error (root mean square error values lower by 8.2-9.4%) and of episodic prediction ability (false alarm rate values lower by 7-13%). The results demonstrate that artificial neural networks (ANNs), if properly trained and formed, can provide adequate solutions to particulate pollution prognostic demands.     

Application of the tracer-aerosol gradient interpretive technique to sulfur attribution for the big bend regional aerosol and visibility observational study

A simple data analysis method called the Tracer-Aerosol Gradient Interpretive Technique (TAGIT) is used to attribute particulate S and SO2 at Big Bend National Park in Texas and nearby areas to local and regional sources. Particulate S at Big Bend is of concern because of its effects on atmospheric visibility. The analysis used particulate S, SO2, and perfluorocarbon tracer data from six 6-hr sampling sites in and near Big Bend National Park. The data were collected in support of the Big Bend Regional Aerosol and Visibility Observational (BRAVO) Study; the field portion was conducted from July through October 1999. Perfluorocarbon tracer was released continuously from a tower at Eagle Pass, TX, approximately 25 km northeast of two large coal-fired power plants (Carbon I and II) in Coahuila, Mexico, and approximately 270 km east-southeast of Big Bend National Park. The perfluorocarbon tracer did not properly represent the location of the emissions from the Carbon power plants for individual 6-hr sampling periods and attributed only 3% of the particulate S and 27% of the SO2 at the 6-hr sites in and near Big Bend to sources represented by the tracer. An alternative approach using SO2 to tag "local" sources such as the Carbon plants attributed 10% of the particulate S and 75% of the SO2 at the 6-hr sites to local sources. Based on these two approaches, most of the regional (65-86%) and a small fraction (19-31%) of the local SO2 was converted to particulate S. The analysis implies that substantial reductions in particulate S at Big Bend National Park cannot be achieved by only reducing emissions from the Carbon power plants; reduction of emissions from many sources over a regional area would be necessary.     

Spatial modeling for air pollution monitoring network design: example of residential woodsmoke

The purpose of this paper is to demonstrate how to develop an air pollution monitoring network to characterize small-area spatial contrasts in ambient air pollution concentrations. Using residential woodburning emissions as our case study, this paper reports on the first three stages of a four-stage protocol to measure, estimate, and validate ambient residential woodsmoke emissions in Vancouver, British Columbia. The first step is to develop an initial winter nighttime woodsmoke emissions surface using inverse-distance weighting of emissions information from consumer woodburning surveys and property assessment data. Second, fireplace density and a compound topographic index based on hydrological flow regimes are used to enhance the emissions surface. Third, the spatial variation of the surface is used in a location-allocation algorithm to design a network of samplers for the woodsmoke tracer compound levoglucosan and fine particulate matter. Measurements at these network sites are then used in the fourth stage of the protocol (not presented here): a mobile sampling campaign aimed at developing a high-resolution surface of woodsmoke concentrations for exposure assignment in health effects studies. Overall the results show that relatively simple data inputs and spatial analysis can be effective in capturing the spatial variability of ambient air pollution emissions and concentrations.     

Improving nitrogen removal using a fuzzy neural network-based control system in the anoxic/oxic process

Due to the inherent complexity, uncertainty, and posterity in operating a biological wastewater treatment process, it is difficult to control nitrogen removal in the biological wastewater treatment process. In order to cope with this problem and perform a cost-effective operation, an integrated neural-fuzzy control system including a fuzzy neural network (FNN) predicted model for forecasting the nitrate concentration of the last anoxic zone and a FNN controller were developed to control the nitrate recirculation flow and realize nitrogen removal in an anoxic/oxic (A/O) process. In order to improve the network performance, a self-learning ability embedded in the FNN model was emphasized for improving the rule extraction performance. The results indicate that reasonable forecasting and control performances had been achieved through the developed control system. The effluent COD, TN, and the operation cost were reduced by about 14, 10.5, and 17 %, respectively.     

Spatial Modeling for Air Pollution Monitoring Network Design: Example of Residential Woodsmoke

Abstract

The purpose of this paper is to demonstrate how to develop an air pollution monitoring network to characterize small-area spatial contrasts in ambient air pollution concentrations. Using residential woodburning emissions as our case study, this paper reports on the first three stages of a four-stage protocol to measure, estimate, and validate ambient residential woodsmoke emissions in Vancouver, British Columbia. The first step is to develop an initial winter nighttime woodsmoke emissions surface using inverse-distance weighting of emissions information from consumer woodburning surveys and property assessment data. Second, fireplace density and a compound topo-graphic index based on hydrological flow regimes are used to enhance the emissions surface. Third, the spatial variation of the surface is used in a location-allocation algorithm to design a network of samplers for the woodsmoke tracer compound levoglucosan and fine particulate matter. Measurements at these network sites are then used in the fourth stage of the protocol (not presented here): a mobile sampling campaign aimed at developing a high-resolution surface of woodsmoke concentrations for exposure assignment in health effects studies. Overall the results show that relatively simple data inputs and spatial analysis can be effective in capturing the spatial variability of ambient air pollution emissions and concentrations.     

Neural Network and Multiple Regression Models for PM10 Prediction in Athens: A Comparative Assessment

Abstract

Particulate atmospheric pollution in urban areas is considered to have significant impact on human health. Therefore, the ability to make accurate predictions of particulate ambient concentrations is important to improve public awareness and air quality management. This study examines the possibility of using neural network methods as tools for daily average particulate matter with aerodynamic diameter <10 µm (PM₁₀) concentration forecasting, providing an alternative to statistical models widely used up to this day. Based on a data inventory, in a fixed central site in Athens, Greece, ranging over a two-year period, and using mainly meteorological variables as inputs, neural network models and multiple linear regression models were developed and evaluated. Comparison statistics used indicate that the neural network approach has an edge over regression models, expressed both in terms of prediction error (root mean square error values lower by 8.2–9.4%) and of episodic prediction ability (false alarm rate values lower by 7–13%). The results demonstrate that artificial neural networks (ANNs), if properly trained and formed, can provide adequate solutions to particulate pollution prognostic demands.