This paper briefly introduces the production status of magnesium thermal reduction process, takes the design of furnace temperature controller for magnesium refining furnace as an example, and analyses the influence of intelligent control on the production process of high energy consumption enterprises, and finally points out the development direction of intelligent control in high energy consumption industries. In response to the national policy, Northwest China severely curbed the continued expansion of high-energy-consuming industries and advocated energy conservation and emission reduction. Many enterprises were involved in the storm of large-scale restructuring due to unqualified environmental protection. Magnesium metal, as one of the most profitable industries in the coal chemical industry, was no exception. High-energy-consuming and high-pollution enterprises were in a dilemma. However, with the increasing development of Northwest China and the abundant coal resources in Northwest China, the coal chemical industry has been vigorously developed in Northwest China in the past few years. Therefore, magnesium project is a hot project in Northwest China. In response to the government’s call to serve the local economy, save enterprise costs and improve production efficiency, our unit has also invested a lot of funds to develop low-energy heating furnaces in cooperation with enterprises. We have successively cooperated with Ningxia Huayuan Metal Magnesium Co., Ltd., Quanshi Magnesium Co., Ltd., Xinda Magnesium Co., Ltd., Inner Mongolia Ordos Metal Magnesium, Shenmu Metal Magnesium in Shaanxi Province and Dafeng in Shanxi Province. Regenerative reduction furnace has been popularized in Ningxia, northern Shaanxi, Inner Mongolia and Shanxi metallurgical magnesium industry after nearly ten years’efforts. The traditional production line heating furnace can hardly be seen. According to the survey, the introduction of this new technology not only greatly reduces the workload of workers, but also further improves the working environment. The production cost per ton of magnesium is 160-240 RMB less than that of the traditional heating furnace. The survey data of each enterprise are different. According to the individual survey, the basic annual production capacity of several large units is more than 12,000 tons.
According to the minimum saving standard, 192,000 yuan per year can be saved for enterprises. Sub-enterprises may reach more than 2 million yuan. These are only the cost savings of furnace body transformation.
In fact, there is a crucial problem in enterprise investigation-control problem. These years, we have been thinking about this problem. Because we only participated in the mechanical and civil construction parts before, one is that we did not have enough knowledge reserve for control, the other is that we did not intentionally do it, but if the control problem is solved. As far as fuel and manpower can be reduced, there are seven workers in one furnace in the reduction workshop and an average of six in the refining workshop according to the production line. If the control problem can be solved, I think that at least two people can be extracted from each furnace in the reduction workshop, at least 14 people can be reduced in the refining workshop, and the annual production capacity of 12,000 tons of enterprises is basically 24 stoves in the reduction workshop and refining workshop, so as to realize self-reliance. After dynamic control, the unit can use at least 80 workers less, and save at least 2.4 million yuan annually. This part is only the wage-level welfare benefits. At present, there is no reliable data to support fuel savings. Because the control of metal magnesium industry in traditional production process is manual control, and the chemical performance of metal magnesium is more active (after metal sodium), and reduction workshop and refining workshop are the most dangerous and difficult to grasp, so the automatic control of metal magnesium reduction workshop and refining workshop may not be well known. The following is a case study of a magnesium refining project designed, programmed and debugged for the refining furnace of the magnesium refining workshop. The following design is made for the automatic control of the system. In order to improve the stability and reliability of the control system, the upper computer of the system uses two industrial computers, one engineer station and one operator station, which are independent of each other and servers. The upper computer is the man-machine interface between the operator and engineer and the system, which enables the relevant personnel to understand the field execution mechanism, relevant data, fault alarm and report file query through the man-machine interface.
It improves the efficiency of production, the authenticity of data, the simplification of workshop control and related fault query.
Because there are too many variables to be monitored and controlled in the field, one rack can not meet the control requirements, so the system expands three racks. The system is mainly divided into furnace temperature control, voltage control and furnace pressure control. All the 4-20MA signals of the field instruments are sent to the PLC by hard wiring, and then the data are transmitted to the host computer through industrial Ethernet after processing. All start-stop control on the upper computer is also transmitted to the PLC by Ethernet communication, which is then reflected to all the executing agencies on the site. The staff monitors and controls the existing equipment on the spot through the upper computer. There are manual control units on the spot. When a upper computer crashes or fails, it will not affect the normal operation of the system. Because the two upper computers are redundant, they are servers of the control system, and the workbench can also be controlled manually.
This also greatly improves the stability and reliability of the control system. The simulation tool used in this paper is MTLAB. MATLAB is a set of high performance computing and visualization software launched by Math works in 1982. It has powerful calculation function, rich and convenient graphics function, wide application range, high programming efficiency, strong expansion ability, simple sentences, easy to learn and use. It integrates numerical analysis, matrix operation, signal processing and graphics display. Its powerful extended functions provide a basis for applications in various fields. In particular, various MATLAB toolboxes developed by experts and scholars in various fields have made MATLAB from an engineering calculation software to a powerful tool for automatic control calculation and simulation. Simulink is one of the toolboxes in MATLAB. Its main function is to realize dynamic system modeling, simulation and analysis, so that the system can be simulated and analyzed beforehand before the actual system is made, and the system can be corrected in real time or the parameters of the control system can be debugged and adjusted according to the best effect of simulation, so as to improve the performance of the system and reduce the design process.
Repeated modification of time to achieve the goal of efficient development of the system. It is simple to use, powerful, and supports continuous, discrete and mixed linear and non-linear systems. Simulink contains many sub-model libraries, and each sub-model library contains many functional modules. And using the mouse dragging, building block editing method, simple, convenient, intuitive and efficient. Users can also write custom function in language, and connect module control program to simulation model through M-function module in Simulink. Fuzzy logic toolbox is a design tool of fuzzy controller which can be embedded in Simulink simulation view. Here I use Simulink to simulate the temperature of resistance furnace. The two-dimensional input and one-dimensional output are used in the fuzzy controller. The input variables are temperature deviation E and temperature deviation change rate EC, and the output is temperature control change U. For each input and output variable, it is normalized to [-1,1] interval, and seven fuzzy sets NL, NM, NS, ZE, PS, PM, PL (where N represents negative and P represents positive) are used. Each input fuzzy set adopts triangular function, and each output fuzzy set adopts single numerical type. The fuzzy rules used in this paper are shown in Table 1-1. Through the above simulation attempts, it can be seen from the simulation graphics that the fuzzy control can quickly stabilize the expected value and oscillate slightly near the expected value, while the PID control can only stabilize to the expected value after a period of time. Although after stabilization, the PID control will stabilize around the expected value while the fuzzy control oscillates around the expected value, it can be clearly seen from the comparison of Fig. 1-7 and Fig. 1-10 that the fuzzy control can also stabilize around the expected value in the case of disturbance, and the influence of disturbance on the PID control is considerable. Therefore, through simulation, we can see that the stability of fuzzy control is far superior to that of PID control when dealing with the industrial resistance furnace, which is uncertain, large inertia, large time delay and non-linear control object. Resistance furnace is a kind of control object with uncertainties, large inertia, large time delay and serious non-linearity. It is difficult to accurately model, and it is difficult to achieve good control effect with traditional PID control. In addition, the control parameters of conventional PID are usually adjusted manually, the adjustment process is tedious and wastes manpower and material resources. Fuzzy control is actually a model-free control method based on expert knowledge. It does not depend on the mathematical model of the object to realize some human intelligence. According to the inherent mechanism of resistance furnace temperature control, considering various factors comprehensively, the temperature control system of resistance furnace is designed by means of mathematical analysis, and it is determined that resistance furnace is a first-order pure lag system.
On the basis of comparing the advantages and disadvantages of fuzzy control and PID control, and according to the accumulated experience of long-term manual operation of resistance furnace temperature control system, the fuzzy control method is applied. This control method combines the characteristics of fuzzy control and conventional PID, and enhances the advantages and avoids the disadvantages. The simulation results under the SIMULINK environment of MATLAB show that for the nonlinear system with pure delay, it has the characteristics of fast response of conventional fuzzy control and high steady-state precision of PID control. In addition, it has strong adaptive ability and robustness, thermostatic element so it can meet the requirements of multi-disturbance, variable parameters and non-linear control process. Moreover, the program is relatively easy to implement. The system realizes many functions, such as real-time monitoring, fault alarm, storage and query of historical data, display and printing of report forms, and the user interface is well operated and easy to learn. The system has good versatility, expansibility and maintainability. Through the above design, the labor intensity of workers can be greatly reduced and the management level can be improved. This design provides an excellent material reference for the automatic design of magnesium refining project. How to further improve the steady-state accuracy of the fuzzy control and apply the advanced fuzzy control strategy to the temperature control of resistance furnace in magnesium industry is the next theoretical research work. The main task of practice is to debug the MCU fuzzy controller accurately.