Advanced control of grain drying process (1)

 Abstract: The development of advanced control in grain drying  process is summarized based on the characteristics of advanced control in this paper. The kernel problems in drying process control are introduced. l is also discussed. Some advices on the development of grain drying process control is proposed.
 Key words : drying;advanced control；adaptive control；model predictive control；expert control；fuzzy control；neural network control 

The basic goal of grain drying is to keep the drying process stable, and obtain the excellent drying quality with low drying cost and energy consumption. Grain drying process is a typical non-linear, multivariable, large lag, parameter correlation coupling unsteady heat and mass transfer process. Grain itself is a complex biochemical substance. In order to achieve the above goal, the drying parameters must be constantly adjusted in the drying process to control the drying process. The automatic control of drying process is an effective means to achieve high quality, high efficiency, low consumption and safe operation. To realize the automatic control of drying process and grain dryer is of great significance to ensure that the moisture content of grain is uniform, the quality of grain after drying, reduce the labor intensity of operators and give full play to the production capacity of the dryer. According to the development goal set by the State Grain Administration in the "Tenth Five Year Plan for the development of science and technology of grain industry", the online monitoring and automatic control of grain drying process has become the key problem to improve the efficiency of grain drying process and an important way to realize the Tenth Five Year Plan. With China's increasing investment in grain depot construction, grain processing industry is increasingly in line with international standards. The automation of grain drying will lay the foundation for China's grain to join the international circulation market.

Features of advanced control

The research on automatic control of grain drying process began in 1960s. At that time, traditional control methods such as feedforward control, feedback control, feedback feedforward control and adaptive control were used. The traditional control theory uses difference equation or transfer function to express the knowledge and existing information of drying process system into analytic formula. However, there are many difficulties in using and designing the control system of grain dryer based on the above control method. The reasons are: (1) the drying process of grain is complex, time-varying and nonlinear; (2) some drying process variables (such as grain quality and color) can not be directly measured, and some variables (such as grain moisture content) may be discontinuous, non-linear and non-linear (3) the process model of the dryer is approximate to the actual process, and needs a lot of calculation time; (4) it is almost impossible to use an appropriate model to represent such a nonlinear, time-varying and time-varying complex system like drying process; (5) there is interaction between the controlled variables and the control variables of the grain dryer; (6) the process model of the grain dryer is similar to the actual process The operating conditions are complex, and the disturbance variable range is wide, which is difficult to control.

Obviously, in order to overcome the above difficulties, it is necessary to improve the traditional control methods of grain dryer, and explore new and more effective control methods. In the 1970s, the progress of electronic industry, especially the development of computer technology, made the so-called advanced control idea widely spread. The goal of advanced control is to solve the control problems of complex industrial processes which are not effective or even uncontrollable by conventional control. In recent years, the rapid development of modern control and artificial intelligence has laid a strong theoretical foundation for the implementation of advanced control system; the popularization of control computer is distributed control system (DCS), and the rapid development of computer network technology provides a powerful hardware and software platform for the application of advanced control. In short, the needs of industrial development, control theory and the development of computer and network technology strongly promote the development of advanced control.

With the rapid development of computer technology, artificial intelligence control theory began to be applied in the control of dryer, which significantly improved the performance of the dryer control system. The traditional control method is not suitable for grain dryer because of its large lag and nonlinear relationship to grain drying process. The progress of artificial intelligence technology is widely used in the field of engineering. Advanced control theory and control method are applied to the automatic control of grain drying process. The control method is constantly improved and the control effect is improved. Since the 1990s, the process control has been developed to intelligence. The intelligent control theory is increasingly combined with the drying technology. The artificial neural network is used to simulate and control the drying process. The system is applied to the prediction of grain quality, the control of drying process and management consultation.

Advanced control system, which is closely related to control theory, instrument, computer, computer communication and network technology, has the following characteristics:

(1) The theoretical basis of advanced control system is mainly model-based control strategies, such as model predictive control. These control strategies make full use of the input and output information of industrial processes to establish system models, and do not need to rely on the in-depth study of reaction mechanism. Recently, knowledge-based control, such as ^ control and fuzzy logic control, is becoming an important development direction of advanced control.

(2) Advanced control systems are usually used to deal with complex and variable process control problems, such as large time delay, multivariable coupling, various constraints between controlled variables and control variables. The advanced control strategy is a dynamic coordinated constraint control based on the conventional single loop control, which can make the control system adapt to the dynamic characteristics and operation requirements of the actual industrial production process.

(3) The implementation of advanced control system needs a high-performance computer as the support platform. Due to the complexity of advanced controller control algorithm and the influence of computer hardware, the advanced control algorithm of complex system is usually implemented on the upper computer. With the continuous enhancement of DCS functions and the development of advanced control technology, some advanced control strategies can be realized in DCS together with basic control loops. The latter method can effectively enhance the reliability, operability and maintainability of advanced control.

2. Development status of advanced control in drying process

The advanced control strategy is the core content of advanced control system. At present, there are many kinds of advanced control strategies. The main advanced control strategies in drying process include: predictive control, fuzzy logic control, neural control, adaptive control and system.

2.1 model based control

2.1.1 adaptive control

The basic principle of adaptive control is to adjust the control parameters at any time according to the changes of drying process parameters and external interference, so as to make the dryer in a good working state. The adaptive control has the advantages of being suitable for a variety of grain dryers, without any data about the dryer's own characteristics, no special requirements for the environmental conditions and grain conditions, the controller's response speed to the disturbance is fast, and the parameters in the control model can be automatically adjusted with the changes of external conditions. Nybrant (1985) of Sweden applied self-tuning technology to the control of cross flow grain dryer. The exhaust temperature of the dryer is taken as the output variable, and the grain discharging rate is used as the controlled variable. The ARMA model is selected to represent the dynamic characteristics of the cross flow dryer. The standard deviation of control error is 0.13 ℃ in the last 50 samples. The results show that the adaptive controller can control the exhaust temperature accurately. Liu Jianjun [5] (2003) studied htj-200 dryer, quantitatively analyzed the system through online sample collection and intelligent optimization algorithm, established the process intelligent model determined by real-time detection data, and then called the artificial intelligence model through intelligent optimization algorithm to obtain the control rules of the system, and the control quantity was given by the control program and output to the execution department after D / a conversion Pieces. Li Xiaobin et al. [3] (1998) studied the advanced control system of vacuum freeze-drying equipment. According to the process requirements of different freeze-drying materials, two adaptive and self-tuning control methods, namely DRA algorithm and critical proportion method, were adopted to solve the lag problem of temperature, the main control parameter of the controlled object.
2.1.2 model predictive control

The new research field of process control theory is model predictive control, which is an optimal control algorithm based on model, rolling implementation and feedback correction. It is especially effective for controlling nonlinear and large lag processes.

Forbes, Jacobson, Rhodes, Sullivan [24] (1984) and eltigani designed a model-based drying controller, whose control behavior is based on a process model and a so-called fake inlet grain moisture content. The drying rate parameters are updated intermittently according to the difference between the predicted value of the model and the moisture content measured at the outlet of the sensor. The difference between Forbes and eltigani controllers lies in the types of process models used in control algorithms. Liu Qiang [25] (2001) of the University of Michigan proposed a model predictive controller for cross flow dryers. The simulation test was carried out on a Zimmerman vt-1210 tower type cross flow grain dryer. The controller established by LabVIEW can operate successfully, and the moisture content of corn at the outlet can be controlled within 0.7% of the set point. The controller can make good compensation for the change of moisture content of grain entering the dryer and the large step change of hot air temperature.

In the study of model predictive control, more work is focused on the establishment and solution of process model, and the drying quality is considered in the model. With the help of partial differential equation (PDEs), P. Dufour [31] et al. (2003) of France extended model predictive control to system model, so that PDEs equation can be applied on a large scale. They proposed a global model to reduce online computing time due to the PDE model based on ^ optimized task solutions. A general MPC framework is developed, which combines with the IMC structure which is widely used in practice. Two feedback loops are used in the imc-mpc architecture to correct the simulation errors caused by process performance and model-based online optimizer. Helge didriksen [29] (2002) of Denmark developed a dynamic first-order law model describing the conversion of mass, energy and momentum of a drum dryer, and applied it to the predictive control of sugar beet drying in sugar mills. The results show that the model has good prediction ability with the change of operation variables and disturbances. Compared with the traditional feedback control and model predictive control, the model predictive control shows better performance. In 1997, I.C. trelea, G. trystram and f. Courtois [27] of France designed a nonlinear predictive optimization control algorithm for batch drying process, which was tested on a pilot scale dryer. Experiments show that the algorithm can deal with important disturbances and failures, and the control algorithm can be easily used in other batch processes, such as freezing, sterilization or fermentation. Some scholars use neural network to model the process of model predictive control. Jay [32] (1996) first applied neural network model to predictive control of drying process. French J.A. Hernandez Perez et al. [33] (2004) proposed a mass and heat transfer prediction model based on artificial neural network. The model takes product shrinkage as a function of moisture, and applies two independent feedforward networks with a hidden layer, in which there are three neural cells, which can predict mass and heat transfer. In the verification of the data device, the simulation is consistent with the experimental kinematic test. The developed model can be used for on-line state estimation and control of drying process.
2.2 intelligent control

Intelligent control is a new theory and technology, it is the advanced stage of traditional control development. This is a kind of control theory which is more close to the mode of thinking of the human brain. It is mainly used to solve the control of complex systems which are difficult to solve by traditional methods. The controller design is free from the bondage of the system model, and the algorithm is simple and robust. At present, intelligent control technologies such as ^ control, neural control and fuzzy control are becoming an important development direction of advanced control.

2.2.1 ^ control

^The system technology can integrate the mathematical algorithm with the operation experience of the control engineer, and make full use of the existing knowledge to achieve the control effect which is difficult to obtain by the traditional control mode. ^The control system runs in a continuous real-time environment. The dynamic characteristics of the system are monitored by real-time information processing, and appropriate control functions are given. The combination of ^ - system technology and grain drying process control can improve grain production efficiency and production efficiency. Liu Mingshan [12] (2001) developed a fuzzy control system for grain drying. The simulation results were compared with the measured data, and they were basically consistent. Liu Shurong [13] (2001) designed a fuzzy control system for high moisture grain drying process by combining ^ - system technology with drying process control. He Yuchun [14] (2001) optimized the drying parameters in the drying process through intelligent control, and worked out the common benefit points of energy consumption, efficiency and quality in the design of drying equipment and the drying process, so that the dryer can dry grains along the common benefit line, so that the equipment is always in good operation in the drying process; at the same time, the temperature measurement and control technology and network technology are interconnected to establish a set of simple and convenient Effective network measurement and control system based on temperature.

2.2.2 neural network control

Neural network can provide an effective method for the modeling of complex nonlinear process, and can be used in process soft sensing and control system design. There are two main applications of neural network in drying process: Modeling and control of drying process.
French J. - L. dirion (1996) [6] and others developed a neural controller to adjust the temperature of a semi batch reactor. The basic experiments formed a learning database of neural networks. The neural controller can provide very good set point tracking and interference elimination. Liu Yaqiu [9] (2000) developed an adaptive PID controller based on single neuron, designed the neural network model of wood drying kiln, described the input and output characteristics of the drying kiln with BP algorithm, and learned and trained the model. Through experiments and simulation, it is proved that the conclusion obtained meets the requirements of error index. Zhang Jili [10] (2003) designed an on-line detection and intelligent predictive control system for grain drying process parameters by combining fuzzy control technology with neural network technology. The results show that the change range of the moisture content in the outlet grain of the dryer under the intelligent control is smaller than that of the manual control, the former is 13.6% - 14.4%, the latter is 12.4% - 14.2%; the fluctuation frequency of the intelligent control is smaller than that of the manual control, the fluctuation period of the former is about 20h, and the latter is about 8h. Wang pin [11] (2003) established the neural network model of the drying tower with the improved BP network algorithm, and established the neural network controller through the neural network model, which realized the intelligent control of grain moisture drying in the arch drying tower system, and improved the quality and efficiency of grain drying.

Liu Yongzhong [8] (1999) applied the theory of artificial neural network system to predict the characteristics of freeze-drying process, and took the drying process characteristic parameters such as drying time, sublimation drying time share, drying product productivity and sublimation interface temperature as the output parameters of the network model, and compared the prediction results of the network with the calculation of the mathematical model, and the predicted results were in good agreement with the calculation results. Zheng Wenli [7] (2000) used artificial neural network to intelligently simulate the weight change of freeze-drying materials during freeze-drying process: the orthogonal experimental results of freeze-drying process conditions were learned, and the learned network was used to predict and optimize the process conditions.

2.2.3 fuzzy control

Fuzzy control is a kind of rule-based control, which directly adopts linguistic control rules. It is based on the control experience or related knowledge of field operators. It is not necessary to establish the mathematical model of the controlled object in the design, so the control mechanism and strategy are easy to accept and understand.

At present, the main application of drying process control at home and abroad is fuzzy control method. Zhang Qin [15] et al. (1994) studied the fuzzy control of the continuous cross flow grain dryer. The operation of the dryer was controlled by adjusting the power of the heater and the rotation speed of the grain unloading auger. The success rate of the experimental control was 86.4%. Li Junming [16] et al. (1996) based on the hot air temperature of drying tower, a skilled operator in corn drying production formulated fuzzy control rules through observation and experience of sensory system, realized speed regulation of displacement motor by using fuzzy control, and proposed that the self-organizing fuzzy controller of cross flow corn dryer should adopt open-loop fuzzy control system to solve the problem of corn drying The problem of large delay in the process. Li Yede and Li Yegang [17] (2001) designed a fuzzy intelligent controller with 89C51 single chip microcomputer as the core. Through the on-line drying test of wheat on a downstream dryer, it was proved that the system has the advantages of short response time, small overshoot and high control accuracy, but the fluctuation of moisture content in the inlet grain will affect the drying process.

Many graduate students are engaged in the research of fuzzy control of grain dryer. Meng Xianpei [18] (2003) of Northeast University used fuzzy set theory and optimization algorithm to establish the intelligent model of grain drying system and the fuzzy rules of fuzzy control system in the intelligent modeling and intelligent control of grain drying tower, and designed the fuzzy controller of the system. Tang Xiaojian of Harbin Institute of technology [20] (2003) studied the multivariable fuzzy control method of mixed flow grain drying tower based on TS model, carried out control simulation of the system, and compared with manual control method and traditional fuzzy control method. Cao Yanming [21] (2000), from South China Agricultural University, developed the automatic control system of rice circulation dryer by using the design method of fuzzy control simulating human thinking mode according to the characteristics of high humidity rice circulation slow drying process. Su Yufeng [23] (2002) of Northwest Institute of light industry adopted fuzzy algorithm based on workers' practical operation experience and used single chip microcomputer to control the freeze-drying system, which improved the automation degree of equipment.