Study On Intelligent Control System For The Pulverizing Process Of Alumina Sintering Rotary Kiln

Although bauxite ore is very rich in China, the alumina-silica ratio of the ore is generally low. Therefore, the alumina sintering process technology has widely been used to produce alumina. The pulverizing system of alumina sintering rotary kiln is utilized to pulverize raw coal into coal powder and make the particle size and moisture content within their required ranges, with the remaining heat produced by cooling clinker. In pulverizing process, qualifications of the particle size of coal powder and moisture content can increase the heat efficiency of combustion and decrease the assumption of coal powder. Moreover, increase of the mill output is very important in terms of improvement in pulverizing efficiency and reduction of electrical energy assumption. Thus, it is of importance and significance for improving the alumina quality and the production efficience to control the particle size and moisture content of the coal powder within their required ranges and in the meanwhile to maximize the mill output.In pulverizing process, the index of particle size of coal powder can not be directly measured online, and has the complicated relations with the mill load, the temperature of hot air, the moisture content of raw coal, the particle size of raw coal and ball loads in mill; relationship among mill loads, the temperature of mill outlet and the speed of coal feeder is of a strong nonlinear nature, which is also affected by the temperature of remaining heat. Moreover, the remaining heat of pulverizing process is low and varies frequently, thus the exsiting control methods are difficult to control the index of particle size and mill load.In industrial process, manual operation on pulverizing process is mainly used. However, on-site operators cannot accurately recognize the operatioanl conditions, and can not dertermin the coal feeding in time. Such a manual control can lead to possible operational faults such as the "mill blockage" and the "powder-enwrapped balls". Moreover, for the manual control, the particle size of the coal powder fluctuates violently as the boundary conditions vary, such that the particle size of coal powder is unqualified and the electric energy assumption is large.The work conducted in this thesis was supported by the National 863 High Technology Program project of "the intelligent optimal control technology and its application of energy assumption index for complex industry process". The objective of this thesis is to control the particle size and moisture content within their required ranges and in the meantime to maximize the mill output. This thesis focuses on the research on the intelligent control system for the pulverizing process of alumina sintering rotary kiln. The achievements are as follows: based on the dynamic character analysis of this system, an intelligent operational control method which consists of an intelligent setting layer of mill load and a process control layer is proposed; an intelligent control system which includes hardware platform, software platform and intelligent control software has been successfully applied in a plant in China. The main research of this paper concludes the following contributions:1. The dynamic characteristics between the particle size of coal powder and mill load, and the dynamic characteristics between the mill load, temperature of mill outlet and the coal feeding are first analyzed separately. Then, an intelligent control method of pulverizing process of alumina sintering rotary kiln has been proposed, which includes a mill load intelligent setting layer and a process control layer. The intelligent setting control method for mill load produces a setpoint of mill load, according to the target of the particle size of coal powder. The process control layer using the mill load switching control method makes the mill load to track its setpoint by adjusting the speed of coal feeder, thus this index can be controlled.(1) An intelligent setting control method for mill load has been presented to make the particle size of coal powder within its required range and to maximize the mill output. This control method includes an intelligent setting model for mill load, a soft measurement model based on the least square-support vector machines (LS-SVMs), and a compensation model based on fuzzy rules. The intelligent setting model consists of a pre-setting model and a compensation model for the variation of iron ball load. This model realizes the setting of the mill load by combining the multi-model technique, the rule based reasoning and the compensation method. The soft measurement model estimates the particle size of the coal powder according to the auxiliary variable using the LS-SVMs method. The feedforward compensation model modifies the setpoint of mill load using the error between the output of soft measurement model and the target of the particle size.(2) An intelligent switching control method of mill load has been proposed to control the mill load and the temperature of mill outlet within their target ranges, and to make the mill operational noise represent the mill load at the optimal setpoint. It consists of a switching mechanism, a coordinate controller for the mill load and the temperature of mill outlet, a PI controller for the mill load, and a rule based overload controller. First, a rule based reasoning method is used to recognize the system conditions. The three following controllers are designed to cope with the three different conditions in a switching manner, respectively. Detailedly, the coordinate controller for mill load and the temperature of mill outlet is implemented under low the temperature condition. The rule based reasoning overload controller is used under the overload condition, and the PI controller is used under the normal condition.2. The intelligent control system has been designed and developed for the pulverizing system of alumina sintering rotary kiln. It includes hardware platform, software platform, and intelligent control softwares. The hardware platform includes the measurement instruments, the transmitters, the actuators, the supervision computer and touch screen, etc. The software platform is based on the configuration software of process control. The softwares including the intelligent setting control software for mill load, the process control software, the switching control softeware for mill load, and monitoring software are designed and developed to implement the above mentioned control methods.3. The intelligent control system has been applied in a pulverizing process of alumina sintering rotary kiln. This application in practice industry shows that the switching control method copes with the automatic control of mill load, and makes the mill load and the temperature of mill outlet within their required ranges, and thus avoids the faults such as "mill blockage" and "powder enwrapped ball" caused by low temperature of coal powder or overload conditions. And the intelligent setting control for mill load makes the particle size within its required range, and the electric energy assumption reduced by 4.56% compared with the manual operation.The successful application has shown that the proposed intelligent operational control method and the designed control system have important and extensive applications for other pulverizing processes. Also, it can be as a reference for the operational control of complex industrial process which can hardly be described using any accurate mathematical models.