Abstract:Coal blending is becoming increasingly common as more and more off-specification coals are received in coal-fired power plants, given the present coal market in China. This situation requires optimization of the operating parameters for matching the varying coal properties. The motivation for such optimization includes confirming good performance of the units regarding the security, the economy and environmental protection. However, the current adjustments to operation of the plant rely mostly on human experience because of the imperfections of existing theoretical models for coal-blend combustion. In this paper, a Case-Based Reasoning (CBR) method providing online decision-making for optimization of coal-blend combustion was investigated using cases representing successful operation of the unit for specific coal blends and loads. A case base containing a wealth of knowledge about optimal operation modes was constructed from a large number of cases. The development process for the CBR system includes case design, case evaluation, case generation, case retrieval and case reasoning. Case evaluation focused mainly on heating surface security, output capability, slagging tendency, comprehensive fuel consumption and pollutant emissions. Five indexes were introduced to quantify the above characteristics based on actual combustion parameters. A case-generating algorithm employing an evolutionary strategy was proposed in which the case base evolves while retaining new cases. Two methods for measuring case similarity - termed entirely similarity and eigenvalue similarity - were used for case retrieval. Run-time optimization strategies were recommended by the case-reasoning model based on the current operating status. The CBR system using Browser/Server framework were successfully applied to a 600-MW power plant, which provided an opportunity for coal-blend combustion optimization. (C) 2014 Elsevier Ltd. All rights reserved.

Bayesian Rough Set and Mixture Experts Model Based CBR System

Design and Implementation of Context-Aware RBAC Model Based on Reasoning

Research on rapid diagnosis algorithm for complex system based on Bayesian theory

Quality problem oriented extension knowledge representation and diagnostic reasoning

Meters Design Expert System Based on Mixed CBR and RBR

Case-based Reasoning for Short-Term Load Forecasting Based on Hierarchy Matching

Research and implementation of a multi-fault diagnosis method based on case-based reasoning

Estimating Machining Processing-hour of Injection Mould Using Case-based Reasoning

Case Reuse Based on Fuzzy Reasoning – Adaption for Emergency Management

An improved case-based reasoning method and its application to predict machining performance

Machine learning based feature selection and knowledge reasoning for CBR system under big data

An Online Case-Based Reasoning System for Coal Blends Combustion Optimization of Thermal Power Plant

Improved Case-based Reasoning for Short-term Load Forecasting

A Methodological Approach to Model CBR-based Systems

Developing A Case-Based Reasoning Model for Safety Accident Pre-Control and Decision Making in the Construction Industry

Multimode process data modeling: A Dirichlet process mixture model based Bayesian robust factor analyzer approach

Mixture Bayesian Regularization of PCR Model and Soft Sensing Application

A Similarity Measure Based on Extension Distance and Its Application in CBR

Scheme Design Based on Case-Based Reasoning in Utility Boiler

Research on an advanced intelligence implementation system for engineering process in industrial field under big data

Research on Application of Clustering Algorithm in CBM