Development and Application of Expert System for Lime Furnace Condition Diagnosis Yi Jinping, Wen Dunwei, Deng Shengxiang, Zhou Naijun, Zhou Yimin (Department of Applied Physics and Thermal Engineering, Central South University, Changsha 410083, China) Various parameters and in-furnace process for detection and off-line detection The results of numerical simulation have developed a real-time furnace condition diagnosis expert system for a Y4mX21m lime shaft furnace in a factory, which has created conditions for automation and optimization control of lime furnace production operation.
1刖 lime shaft furnace is the main equipment for lime production and carbon dioxide gas at the same time. The common production process is: limestone and coke are crushed and mixed with a certain ratio (called “cokingâ€) and then added to the furnace from the top of the furnace. In the inside, the air blown in from the furnace bottom flows and heats in the furnace in a fixed bed manner to cause combustion and decomposition reaction. After more than ten hours of calcination, the main product lime is discharged from the bottom of the furnace, and the by-product CO2 is discharged from the top of the furnace. For mechanized lime shaft furnaces, this production process is continuous. Although the lime production process is relatively simple, the system is controlled and designed by computer.
Single, but the chemical reaction, flow and heat transfer process in the furnace are strongly coupled, so the lime furnace is a typical complex control object with large hysteresis, multivariable, nonlinear, significant uncertain factors, etc., and the conventional control strategy is difficult. Achieve control objectives. The automatic control of the furnace can stabilize the decomposition rate and CO2 concentration of limestone, increase production and reduce energy consumption. In this study, a *4.0mX (diameter X high) lime furnace in an alumina plant was used as the object, and the field measurable information was used to build and analyze the process in the furnace. The fault type and the corresponding treatment countermeasures, the application of the expert system theory to develop the online diagnostic expert system of the furnace condition, and lay a foundation for the next step to establish a fuzzy intelligent control system to ensure the operation optimization of the lime furnace and realize the automation of production.
2In-furnace process online inspection and computer simulation analysis 2.1 Basic work In order to finally realize the automatic operation of lime furnace, the automatic measurement of raw materials, fuel and air volume and computer remote operating system were first developed, and the exhaust gas temperature (top temperature) gray was realized. Continuous automatic detection of temperature, CO2 concentration and multi-point furnace inner wall temperature. These measurable data, together with data on the raw materials and fuel components obtained from off-line testing, create conditions for the simulation calculation of the reaction and heat transfer processes in the furnace. In order to make the monitoring process more reliable, a detection device such as oxygen concentration, air temperature and ambient temperature is set up to form a more complete detection system.
2.2 Simulation analysis of furnace process The simulation analysis result of furnace process is an important basis for furnace condition diagnosis and control decision. This study focuses on the limestone decomposition rate and raw material pre-tropical and calcination in the furnace which has important guiding significance for production. The "three-band" position of the reaction zone and the product cooling zone, as well as the thermal engineering results, can also record and display the historical curve of the simulation results.
1 On-line monitoring of calcination rate of limestone in traditional production In the traditional production, the calcination rate of limestone is often calculated by offline sampling analysis and test data, and the hysteresis is large, which makes it difficult to timely and accurately reflect the calcination condition, so the operator can not take active and effective in time. Measures to adjust the operation. According to the heat balance and material measurement, the on-line analysis model of limestone decomposition rate in the furnace is derived, that is, the top temperature obtained by on-line detection under the premise that the raw materials and fuel components are certain, the feeding amount, the coke ratio and the air volume are known (on-line detection). , ash temperature and dioxide concentration data, etc., through the degree calculation to get "real time"
Decomposition rate.
2In-furnace temperature distribution and “three-band†height online numerical simulation divides the total height of the material layer into several sections, and takes a section as the control body by mass conservation and compositional equilibrium equation, and derives the output side of each control body according to the energy balance. Calculate the equations of the material temperature and gas temperature, and set: the pre-tropical zone ends when the outlet material temperature reaches 900 * C, and the reaction zone (calcining zone and combustion zone) is calculated until the decomposition rate and the mechanical incomplete combustion rate reach the set value. until. Using the computer to analyze the obtained equations, the "three-band" position (ie height) and the furnace temperature distribution can be calculated and displayed online.
3 Furnace Diagnostic Expert System This expert system adopts an extended production structure consisting of two libraries and three subsystems, for example.
Figure i furnace condition diagnosis expert system structure diagram dynamic database for storing lime furnace operating process parameters (such as furnace top exhaust temperature, ash temperature, multi-point furnace wall temperature, air volume and blast pressure, C2 concentration, blanking Volume and coke ratio, layer height, oxygen concentration, air temperature and ambient temperature, etc., simulation results of furnace process (such as limestone decomposition rate, pre-tropical zone, calcined zone and cooling zone "three belts" height, The temperature distribution in the furnace, etc.) and the conclusions inferred by the expert system (real-time diagnostic results of the furnace conditions) are also stored in the database.
The knowledge stock is based on the expert knowledge expressed by the production rules, including three judgment criteria for the diagnosis of the furnace condition and 15 fault judgments and operational countermeasure rules. The on-site technicians can delete or modify these judgment rules at any time; the consultation subsystem The function of the inference engine is to call the data in the dynamic database and the knowledge in the knowledge base to infer the operation status of the lime furnace according to a certain inference strategy and select appropriate operational countermeasures; explain the subsystem output inference conclusion and answer the operator's inquiry. Etc.; Knowledge acquisition subsystem is the interface between the system and the expert, completing the task of learning from experts and machine learning.
3.1 Dynamic database detection system The three types of data such as the operating parameters of the lime furnace, the numerical simulation results of the furnace process and the data knowledge introduced by the system are described by the zero (ie object-attribute-value) triplet mode. For example: (The top temperature system 160 contains 8 different types of objects, namely: 1 material layer height, 2 flue gas components, 3 lime mass, 4 furnace top parameters, 5 calcination conditions, 6 fault alarms, 7 batching conditions, 8 operational countermeasures. Each object is characterized by several attributes, each of which has a credibility factor CF (the value range of CF is 1~+1). Indicates how much the system trusts the value of this attribute.
Each object is stored with all the attributes and values ​​that describe it, forming a complete description of the object. Each object in the database is ordered and related to each other. According to the relationship between them, it can be organized into a tree structure. There are three types of nodes on the tree, and all data describing an object is stored on the value node.
32 Knowledge acquisition and representation The expert knowledge of the system is provided by the workshop technicians based on the field experience, and is collected and collected by the system developer. Combined with the lime furnace production process analysis, the operation experience, furnace failure and treatment methods are formed. The fault judgment rule and the three furnace condition judgment criteria (corresponding to three furnace conditions: good, better, and poorer respectively) are mainly expressed by production rules, and each rule has a “Rule name and is *IF Form representation of ...THEN...". Field technicians can modify and expand the rules.
The behavior part of the rule is represented by a behavior function that specifically represents the action. The system uses three behavioral functions: CONCLUDE, CONLIST and TRANSList. Among them, one form of CONCLUDE function is: CONCLUDECPVTALLYCF, C, P, V represent context, parameters and values ​​respectively, and TALLY stores variables that trust the degree of rules. CF is the credibility factor of the rule. When the value of TALLY is greater than 0.2, if CONCLUDE is in poor condition, use the fault judgment rule to find out the cause; determine the countermeasure to eliminate the furnace operation failure; choose a suitable operation mode.
The first two steps are the process of diagnosis of the furnace condition, and the second two steps are the process of finding countermeasures for the operation of the furnace.
IF has an operational failure and has considered other possible potential failures, even if it has not been alarmed; THEN compiles objects that can effectively troubleshoot based on sensitive data, or selects the best countermeasure from the compiled list; ELSE indicates operation Normal and automatic operation of the furnace.
Once a possible fault is identified, the best countermeasure for troubleshooting is formed. The target rule is to trigger an action to generate a possible troubleshooting list and select an optimal countermeasure from which to guide the operator to select a reasonable amount of material to be fed. Compared with the amount of blasting, timely feeding and ashing.
The composition and application function of the 4 furnace condition diagnosis system are called.
The reasoning of the control strategy knowledge is reverse reasoning and depth first search. The reasoning steps are as follows: 1. Read the data in the dynamic database, and judge whether the furnace condition is normal or not according to the judgment rules stored in the knowledge base; 4.1 System integration and model The interrogation limestone furnace diagnostic expert system is integrated with the on-line detection and display of the furnace operating parameters, simulation calculation and result display, but it is relatively independent from the automatic detection system software, and has a strong flexibility. For safe operation and easy maintenance. It is the system composition diagram.
Lime furnace online simulation and furnace condition diagnosis system composition Figure C. Minimize the small amount of fire, to ensure that the temperature can be maintained, so that the furnace is not overheated. In order to ensure the reliability and stability of ignition and small fire, the small fire gas solenoid valve and ignition solenoid valve are interlocked to ensure safe and reliable ignition, and it is not easy to be blown out by strong wind and large gas after ignition.
After adopting the above measures, the furnace ignition production was successful once, and the temperature uniformity reached *2*C, which was greatly improved compared with the previous one, and the energy consumption was also greatly reduced. The working environment of the workers was significantly improved, the factory was extremely satisfied, and helped me. The hospital promoted this type of combustion temperature control in the non-ferrous industry.
3 Conclusion At present, the furnace has been running for one and a half years, the production is basically good, and can adapt to the pressure fluctuations of its natural gas pipeline network.
The problem found was mainly that the temperature at the bottom of the burner was too high, and the temperature at the top of the furnace was too high, which easily caused damage to the electrical components. After discussing with the factory, the problem was solved by improving the internal structure of the burner and repairing the thermal short circuit at the top of the furnace.
In short, the high-speed temperature control burner and temperature control system are safe and reliable in application in the heat treatment high temperature furnace, and the application effect on the medium and low temperature furnace in the nonferrous industry is also very good, which is worthy of promotion.
The side table is one of the main products we produce in our factory, with a variety of different designs and solid structures, you will love our products, all our products are designed in-house, if you have a style you like you can also send us pictures of the design. We can also do it for you and look forward to working with you.
Our factory mainly develops and produces various types of disassembled furniture, simple furniture, storage boxes,Our products include: multifunctional storage rack, assembled Coffee Table, side table, computer table, desk, chair, dining table, 3-layer shelf, 4-layer shelf, 5-layer shelf, wall shelf, tissue tube, storage basket, carrying basket,laundry rack ,etc .
Modern decoration is all about simplicity, cleanliness and generosity, and as our lives become more and more fast-paced, we have a lot of things that need to be stored and categorised to make our homes look more tidy and convenient.
Side Tables,Working Desk,Side Table Storage,Side Table Stand
Jiangmen Lihua Import & Export Trading Co., Ltd. , https://www.jmlihua.com