In the production process of aluminum sheet, strip and foil, in order to reduce the friction coefficient, reduce the rolling pressure, cool the rolls, control the roll shape, and improve the surface quality, shape and thickness accuracy of the strip, it is necessary to spray lubricating coolant (i.e. rolling oil) on the rolls and roll gap [1]. Rolling oil is one of the main consumable auxiliary materials in the production process of aluminum strip and foil, and it can be recycled. A certain rolling oil company in Shanghai has developed an aluminum foil rolling oil recovery system through the introduction of foreign technology and independent redevelopment. The recovery system absorbs the evaporated rolling oil through washing oil, and then a centrifugal pump provides power for the circulation subsystem. The centrifugal pump has the characteristics of reliability, easy adjustment and stable and uniform flow.
During the commissioning and operation of the rolling oil recovery system, it was found that the working state of the centrifugal pump in the circulation subsystem was affected by many factors. When calculating the working efficiency of the centrifugal pump using similarity theory, some unnecessary factors in the model analysis could be ignored. However, in the rolling oil recovery system presented in this paper, the stable flow rate of the circulation subsystem is a prerequisite for high-quality and high-efficiency recovery of the rolling oil. Therefore, traditional similarity theory is no longer suitable for the control requirements of this system. To solve this problem, this paper attempts to use the least squares method to identify the centrifugal pump in the circulation subsystem, obtaining more accurate model parameters and providing a more reliable control basis. This rolling oil recovery system uses a Siemens S7-300 PLC and an ACS510 frequency converter as the controller for the flow pump. The PLC adjusts the parameters of the frequency converter via the PROFIBUS-DP bus, achieving good control results. 1. Mathematical Model of the Centrifugal Pump The effective output power of the pump refers to the effective energy obtained by the liquid delivered by the pump per unit time. The formula for the effective output power of the pump is:In this system, the flow pump and motor are directly connected, so the formula for the pump's shaft power is:
In equations 1-1 and 1-2, is the fluid density; g is the free-fall acceleration; is the volume of liquid output by the pump per unit time; H is the pump head; and is the pump efficiency.
The pump's working efficiency can be obtained from equations 1-1 and 1-2, i.e.: The relationship curve between flow rate and efficiency is shown in Figure 1. The solid line is the theoretical curve, and the dashed line is the actual curve. According to the principle of motors, the electromagnetic power of an asynchronous motor is [2] : Ignoring the influence of factors such as mechanical losses, the load torque of different types can be approximately expressed as: When the load is a fan or pump, i.e., the torque is proportional to the square of the speed, the mechanical power output by the motor can be obtained as: According to the principles of electrical machinery, angular velocity ω is directly proportional to motor speed, which in turn is directly proportional to the frequency of the power supply output by the inverter. Therefore: If the motor slip does not change with operating conditions, then: From equation 1-8, it can be concluded that, without considering minor factors in the system, the flow rate and speed are not strictly proportional, but rather related to the pump's efficiency. That is, changes in the system's structure during operation will also affect the relationship between pump flow rate and motor speed. Therefore, directly applying similarity theory to analyze the system's operating characteristics cannot achieve precise system control.2. Model Identification Based on Least Squares Method
The centrifugal pump used in the circulation subsystem of this rolling oil recovery system has a rated flow rate of 6 m³ /h, or 100 L/min, and the rated operating frequency of the directly connected motor is 50 Hz. Under the premise that the overall pipeline characteristics of the rolling oil recovery system remain unchanged, the operating frequency of the flow pump is varied according to the direct-connected motor by adjusting the operating frequency of the field inverter. The corresponding data of pump flow rate and motor operating frequency are shown in Table 1.
The data measured in Table 1 were graphically processed using Matlab software, revealing that the similarity theory is not ideal in direct application. Figure 2 shows the straight line representing the relationship between the centrifugal pump's operating frequency and flow rate under the similarity theory, while the scatter plot represents the actual measured relationship between the centrifugal pump's operating frequency and flow rate, indicating a significant error between the two.
Figure 3 shows the relationship curve. From Figure 3, it can be analyzed that the relationship between the operating frequency of the flow pump and the actual flow rate is approximately linear.
Therefore, the least squares method can be used to fit the linear function relationship of the flow pump model, and the identification result is:
Error analysis was performed on the above formula using the standard deviation formula:
The standard deviation of the model is 0.0057, so the identification result is relatively accurate and can correctly describe the controlled object.
3. PROFIBUS-DP closed-loop control of centrifugal pumps
The rolling oil recycling system uses ABB's ACS510 frequency converter and Siemens' 14-inch touch screen. It communicates with the CPU of the S7-300 PLC via the PFROBUS bus to realize closed-loop PID control. The speed of the flow pump is adjusted by calculating the difference between the feedback flow value and the given value through the linear function relationship between the motor speed and the flow obtained by the above flow output model. The frequency of the motor speed is given by the ACS510 frequency converter. The working state of the frequency converter is determined by the control program in the S7-300 CPU module. The entire flow control can be divided into three steps: setting the parameters of the frequency converter; calling the PID instruction block FB41 in STEP7; and setting the PROFIBUS communication parameters [3]. In the Step7 programming software, if PID control is to be realized, the FB41 function block integrated in the programming environment, i.e., the continuous controller function block, can be directly called. When performing multiple PID control, multiple FB41 background data blocks can be established to store the control data of different PIDs respectively [4]. As shown in Figure 4, the FB41 function block contains rich setting and output interfaces, including the setting of the P, I, and D parameters of the PID, the scan cycle of the PID program, and the enable control of the PID parameters. The final adjusted output value, i.e., the motor frequency setpoint, is output from the LMN interface to a user-defined address, and then transmitted from the PLC analog output module to the ACS510 frequency converter via PROFIBUS-DP, realizing closed-loop control. 4. Conclusion This paper mainly focuses on the application of a centrifugal pump in a rolling oil recovery system within a circulating subsystem, studying the working characteristics of the centrifugal pump. The control parameter model is analyzed through similarity theory and system identification. Comparison shows that the model obtained through system identification is more accurate, reducing the error generated by similarity theory in the operation of the centrifugal pump. Finally, closed-loop control of the centrifugal pump by the ACS510 frequency converter is achieved through PROFIBUS-DP bus communication, fulfilling the control requirements of the rolling oil recovery system for a stable flow rate of wash oil. References: [1] Xiao Cuiping, Li Yonggang. Overview of the development of oil mist control technology in rolling mill [J]. Nonferrous Metals Processing, 2005, 34(6): 21-23 [2] Chen Boshi. Automatic control system for electric drive [M]. Beijing: Machinery Industry Press, 2008: 147-157 [3] Li Xiaohai, Nan Xinyuan. Research on the application of PROFIBUS-DP in rural constant pressure water supply [J]. China Rural Water Resources and Hydropower, 2011, 12: 78-90 [4] SIMATIC S7-300 PLC manual [M]. Siemens, 2003.Author Basic Information
Name: Xu Wei Gender: Male Date of Birth: August 1985 Affiliation and Position: Graduate Student, School of Electrical Engineering, Xinjiang University Research Interests: Control Engineering Email: [email protected] Mobile Phone: 14799327226 Contact Address: 2011 Graduate Students, School of Electrical Engineering, South Campus, Xinjiang University, Yan'an Road, Tianshan District, Urumqi, Xinjiang, China Postcode: 830000
