Fuzzy Control and Neural Network Control of Limekiln Process
暂无分享,去创建一个
Traditional control methods have poor performances when applied to industrial processes whose models are strongly non-linear and multivariable-based. Better results can be obtained by applying modern control techniques. In the present paper, emerging techniques like fuzzy logic control and the neural network approach to the control of an industrial limekiln have been reviewed. Control of the kiln process is a demanding task and therefore most of the kilns have been and are still operated without supervisory-level control system. The absence of closed loop controls, however, results in inefficiencies in fuel consumption and variation in reburned lime quality. Furthermore, manual control increases the risk of environmental impacts, especially in the form of reduced- sulfur emissions and the probability of equipment failures such as damage to the refractory linings of the kiln. The multivariable nature and difficult dynamics associated with the kiln process requires that an automation system employ techniques such as decoupling, dead time compensation, dynamic feed forward, optimization, and constraint compensation in order to accomplish the stated operational objectives. The goal of a properly designed kiln control solution should be to implement these techniques in a fashion that is usable and maintainable by mill operations. It is only then that the economic benefits from the control optimization can be sustained long-term. 2. PROCESS DESCRIPTION The rotary lime kiln is a vital piece of equipment in the causticizing process in the pulp and paper mill. Its operational objective is to convert lime mud (calcium carbonate) into calcium oxide (quick lime) that can be further used in the causticizing process in order to produce white liquor. The basic purpose of the causticizing system is to convert green liquor containing sodium carbonate (Na 2 CO 3 ) into white liquor (NaOH) that is used in the kraft pulping process. The lime kiln is a long direct contact heat exchanger. Heat energy is supplied by the combustion of fuel that enters with the primary air at the hot-end of the kiln. The energy is transferred to the lime mud that enters at the cold-end of the kiln, and then flows down countercurrent to the combustion gases due to the kiln inclination and rotation speed. During lime kiln operation, calcination of the incoming lime mud occurs, thus regenerating the hot lime required for the causticizing reaction. Maintaining good control of the kiln temperature profiles helps ensure consistently high lime quality availability, high reactivity, and uniform lime pellet size. This provides the causticizing process with the ability to produce excellent quality white liquor in the most efficient manner. 3. CLASSICAL LIME KILN CONTROL USING PID
[1] Dr. Hans Hellendoorn,et al. An Introduction to Fuzzy Control , 1996, Springer Berlin Heidelberg.
[2] Gary A. Smook,et al. Handbook for Pulp and Paper Technologists , 1982 .
[3] A. Dourado,et al. A neural network based control of a simulated industrial lime kiln , 1993, Proceedings of 1993 International Conference on Neural Networks (IJCNN-93-Nagoya, Japan).
[4] Urpo Launonen,et al. Lime Kiln Optimization: Managing the Inputs to Stabilize the Outcome , 2005 .