Research on Computer Control System of Automobile Variable Section Leaf Spring Rolling Mill

frozenviolet

Research on Computer Control System of Automobile Variable Section Leaf Spring Rolling Mill [Copy link]

Abstract: This paper introduces the main hardware structure and design method of the computer control system of the two-roller automobile variable-section leaf spring rolling mill, as well as the application of the multi-modal intelligent control algorithm in the hydraulic servo system. The control system has a reasonable design, stable operation and high control accuracy.

Keywords: variable cross-section rolling mill, leaf spring, multi-mode intelligent control

The variable section rolling mill is a key equipment for producing automobile variable section leaf springs. It compresses the spring steel plate heated to 950℃ and pulls it through the rotating roller compression and drawing mechanism, so that its cross section changes according to the designed deformation curve. At present, domestic leaf spring manufacturers either use imported rolling mills such as some products of the British Seale Company and the German Poker Company, or use domestic three-roll rolling mills imitating Seale Company and some simple rolling mills. The products produced by simple rolling mills have low precision and large width, and the domestic three-roll rolling mills have large energy consumption and equipment volume.

The two-roller variable cross-section rolling mill is a high-tech rolling mill product currently developed in China. The rolling mill has a compact mechanical and hydraulic structure design, low energy consumption, and no side rollers in the mechanical structure to limit the width of the leaf spring during the rolling process. The computer controls the balance of the pressing amount and tension during rolling and the friction force of the rollers, as well as the position servo control of the pressing and drawing device, to ensure that the width of the leaf spring after rolling is within 1%.

This paper introduces the design of computer control system for two-roller rolling mill. The rolling mill is driven by hydraulic servo system and controlled by IBM-PC industrial control computer and programmable controller.

1. Rolling mill structure and main technical indicators

1.1 Rolling mill structure

Figure 1 is a schematic diagram of a two-roll variable cross-section mill. The mechanical structure of the mill consists of a steel frame support, a roll drive mechanism and a drawing mechanism. The rolls are driven by hydraulic motors, and the upper rolls are controlled by servo cylinders to move up and down. The drawing mechanism consists of guide columns and hydraulic clamping devices, and is controlled by horizontal servo cylinders to move horizontally. The pressure is stabilized by a hydraulic accumulator, and the torque output by the hydraulic motor is controlled by a proportional pressure reducing valve. Two rotary encoders detect the displacement and speed in the X and Y directions.

1.2 Main performance indicators of rolling mill

Maximum rolling length: 1200mm

Minimum rolling length: 300mm

Maximum slab thickness: 50mm

Maximum slab width: 160mm

Maximum rolling force: 760 kN

Maximum pull-out force: 200 kN

Maximum rolling speed: 30m/min

Product thickness tolerance: ±0.1mm

Product width tolerance: Plate width ≤ 100mm ± 1mm, Plate width ＞ 100mm ± 1%

2 Computer control system design

2.1 Overall Design

The controlled quantity of the rolling mill control system can be divided into analog quantity part and switch quantity part. The analog quantity part consists of four channels, which respectively control the output flow of the hydraulic system, the output torque of the hydraulic motor, the movement speed of the drawing cylinder, and the movement speed of the pressing cylinder. The analog quantity part is controlled by an IBM-PC 486DX2-66 industrial computer. The switch quantity part is controlled by a MASTERK200 programmable controller (PLC). These controls include oil temperature control of the hydraulic station, cooling of the rolling rolls, operation button station, brake limit in the drawing direction, lubrication control of the rolling rolls and hydraulic system, state detection and fault alarm of the rolling mill, and interlocking control between related parts. The industrial computer and the PLC communicate through the switch quantity interface. This reduces the amount of calculation of the IBM-PC and shortens the control cycle; avoids the concentration of control risks caused by centralized control, and improves the reliability of the system. Regardless of which part of the IBM-PC or PLC is abnormal, the system can take necessary protection measures.

2.2 Control system hardware structure

The structure of the computer control system of the rolling mill is shown in Figure 2. It consists of a computer, four 12-bit optical isolation D/A interfaces, 64 optical isolation switch input and output interfaces (DI/DO), and a programmable controller (PLC). The computer communicates with the PLC through the DI/DO interface and the RS232 interface. Two 12-bit absolute photoelectric encoders are used to detect the position and speed in the X and Y directions. The oil temperature control part is used to control the oil temperature of the hydraulic station within the allowable range; the roll cooling part controls the cooling of the rolls; the button station is used to input control commands; the brake limit part automatically stops when the drawing mechanism of the rolling mill exceeds the parking range; the lubrication control part controls the lubrication oil cylinder to automatically add lubricating oil to each lubrication point according to the number of operations of the drawing mechanism; the status indication part displays the operating status of the rolling mill; the fault alarm part gives an audible and visual alarm when the rolling mill is abnormal.

2.3 Multivariate Processing

In addition to the position servo control of the drawing and pressing mechanisms during the rolling process, the flow of the hydraulic system and the output torque of the hydraulic motor must be kept constant. The control system of the rolling mill is a multivariable control system.

The servo valve used in the hydraulic system has a closed-loop control link for the valve core position, which can be regarded as a proportional link. In order to simplify the controller structure, according to the design requirements, the drawing mechanism, the flow of the hydraulic system, and the output torque of the hydraulic motor are respectively open-loop controlled, and the system is approximately a single variable system. Since the slope of the cross-sectional deformation curve of the product is small, that is, the movement speed of the pressing device is much smaller than the movement speed of the drawing mechanism. Therefore, even if the drawing direction is open-loop controlled, the desired control effect can be obtained.

2.4 Control Algorithm

There have been successful reports on the use of intelligent control algorithms for electro-hydraulic servo systems [1]. For electro-hydraulic servo systems with dual coordinates such as rolling mills, multi-modal human-like intelligent control algorithms have also been selected. The hydraulic servo system of the rolling mill has a large flow, high pressure, and the maximum speed of the drawing direction is 30m/min. The control object has certain nonlinearity and time-varying properties. When designing an effective control algorithm, the real-time performance of the control system should be considered first. This requires that the structure of the control system should not be very complex, and its decision-making and reasoning should be fast.

Multimodal control is a control algorithm in which the controller adopts different control strategies and control modes according to the different characteristic states of the system during the control process. Characteristic state refers to a set of characteristic quantities that reflect the characteristics, characteristic changes and state of the system, represented by Gi. All characteristic states constitute a characteristic state set, represented by G. Then there is

G＝＜G1, G2,…,Gm＞

Where Gi＝（Gi1，Gi2，…，Gih） （i＝１，２，…，ｍ）

The control mode is represented by Ai, and all control modes constitute the control mode set of the multi-modal controller, represented by A.

A＝＜A1, A2, …, Am＞

where Ai=U=f(Gij) (i=1,2,…,m; j=1,2,…,h)

The realization of control is the realization of the reasoning process from G to A. It can be expressed by the production rule IF G THEN A.

(1) G1: (｜Ｅ｜≤b1)∩(｜｜≤b 2 )

A1: (U(n)=U(n1))　

(2) G2: (EE≥0)∩[(｜E｜＞ｂ)1∪(｜E｜＞b 2 )) ?

A2: U(n)＝U 0＋K p (E+K dＴd E＋Ei / T i E iＴ

(3) G3:(EE＜0)∩[(｜E｜＞b1)∪(｜E｜＞b2)]

A3: U(n)＝U0+KKp(Em(1)+EI+KdTdE)　

In order to obtain good real-time performance and also consider the requirements of control accuracy, as few characteristic states as possible should be selected. Considering the stability of the control system, the error and its first-order derivative are selected as the basic characteristic quantities. After such processing and a large amount of human-machine online learning, the following three basic characteristic states and corresponding control modes are summarized.

Where U(n) is the nth output value of the control quantity; U0 is the holding value of the control quantity when the error passes through zero; b1, b2 are constants; Kp and K are the proportional gain and suppression coefficient respectively; Kd, Ki, Td, Ti are the differential and integral suppression coefficients and time constant respectively; T is the sampling interval; Em(1) is the first error peak value; EI is the interval integral value under the G2 state.

Reference [3] gives the model of the control object and the simulation experimental results of the human-like intelligent control algorithm and compares the control effect with the PID algorithm. From the analysis of the experimental results, the use of intelligent control algorithm to solve the real-time control problem of the hydraulic servo system can achieve better control effect than the PID control algorithm; the design of the controller does not rely on the mathematical model of the object, is insensitive to changes in the hydraulic system parameters, and has strong robustness; the algorithm is simple and the control system has good real-time performance.

3 Software Design

The control software runs under the DOS operating system and adopts C++ language modular design. It is divided into modules such as system setting, automatic rolling, manual operation, zero point debugging, process file editing and modification, file operation, display, statistics, and printing. The Chinese graphical interface is easy to operate. You only need to enter the rolling path to generate the process file. The rolling mill status and rolling mill motion curve are displayed in real time. By modifying the system interrupt 8, multi-task processing of sampling, control, and display can be realized.

In short, the use of multi-modal intelligent control algorithm for the control of the electro-hydraulic servo system can achieve better control effects. It has the advantages of flexible control strategy and simple algorithm. It not only has strong servo tracking ability, but also is insensitive to changes in system parameters, has strong robustness and good control effect. The rolling mill was successfully trial-produced in December 1997 and put into formal production. The rolling speed, accuracy, and operating reliability of the rolling mill all meet the design requirements.