1、第 25 页 共25 页 桂林电子科技大学毕业设计论文 Distributed Cement Plant Control-The IntelligentApproachD. M. STEELMAN, MEMBER, IEEE Abstract Rapidly evolving technology has made a broad variety of control system solutions available to industry. Most recently,microprocessors have formed the leading edge of the technolo
2、gy growth. The capability inherent in the present generation microprocessors will not only change the appearance of control equipment, it will in fact change the methods for implementing plant-wide control. The use of microprocessors in a distributed control mode is discussed. The system configurati
3、on consists of a central processing unit, intelligent operators panel and intelligent distributed field control multiplexers. Each device is described from an operational viewpoint. Special features are presented relative to the system and its applicability for control of a cement plant. In conclusi
4、on, intelligent distributed control offer the way to improved control, greater reliability, and lower installed costs. DISTRIBUTED PROCESS CONTROL First of all, what is distributed control? It is the application of digital technology to a recycled old, but good,idea. For the more technically oriente
5、d, a better definition would be, distributed control is a series of microprocessors strategically located throughout a manufacturing facility and dedicated to performing sequential and process control functions for a specific control area within the facility. Neither of those definitions come from W
6、ebsters dictionary, but they do illustrate the idea of distributed control. With the idea of distributed control now in mind, let us look back 25 years. Then, practically all cement plants were operated using the principle of distributed control. Operators were strategically located throughout the p
7、lant and each had the responsibility for controlling a specific area, such as crushing, grinding, firing, etc. Each operator had at his disposal the necessary equipment to perform the required control functions. Further, the operator was required to supply the necessary intelligence to control his a
8、rea of responsibility. Communications were carried out with the other operators in order to coordinate the overall operation. Upward communications fed information to plant management. At the higher level, decisions were made to improve the coordination further and relate plant activities to market
9、conditions and other key factors. Since then, the cement industry has evolved through several series of automation stages. The first step was to centralize control functions into one large control room. This was brought about by the availability of minature electronic control equipment. More efficie
10、nt plant operation was achieved. Also, we began to understand more about the process itself. The increased process knowledge gained through data analysis produced some interesting para-metric cases. Enter then the digital computer. In order to solve the unique parametric cases, quite complex in natu
11、re, required the problem solving capability of a digital computer. Initial attempts at computer control produced varying de-grees of success. We first used the concept of supervisory control-referred to many times as digitally directed analog controls (DDAC) (Fig. 1). Supervisory control utilized a
12、digital computer superimposed on top of the analog control system 1.Based on the results of solving the process equations, the computer would calculate new setpoints for the individual analog controllers.It was a very costly approach which had some fundamental limitations, most of which were imposed
13、 by the analog con-trollers.The most successful supervisory systems were applied to raw mill blending control. To minimize the limitations of DDAC and make better use of the new found knowledge, a trend toward direct digital control (DDC) began. In a direct digital control system (Fig. 2), the proce
14、ss computer assumes the responsibility of an analog controller1. When the computer detects a deviation between the measured input and desired setpoint, it computes a control equation (algorithm) and transmits a correction signal to the appropriate control valve. Note that in this case it is not a se
15、tpoint to an analog controller. The control equation used to compute the output signal is a form of the standard proportional integral-derivative (PID) algorithm. However,DDC also has its limitations. It must perform the calculation on a repetitive basis for each loop of control. As we increase the
16、workload on a computer because of expanding system size or adding more complexity to the process model, DDC becomes less efficient. In other words, the computer has too much work to do. Several choices exist to resolve this dilemma. The central memory of the computer might be expanded a duplicate system installed to divide the workload or a distributed processing network implemented with several micro-processors located in the field and con
