A control method to overcome valve nonlinearity

Abstract

The present invention discloses a control method for overcoming valve nonlinearity. The method can handle various valve nonlinearity problems including "dead zone", "variable gain", "hysteresis", "hysteresis loop", "gap" and "viscosity", and is used to improve the performance of the control system when the valve nonlinearity is serious. The steps of the method are as follows: establish a valve nonlinear model from valve opening instruction to valve opening feedback; construct a sub-loop control system for controlling the valve opening feedback soft measurement signal, and adjust the parameters of the sub-loop controller so that the control performance meets the requirements; connect the sub-loop to the main loop in the form of cascade control, and adjust the parameters of the main loop controller so that the control performance of the controlled variable meets the requirements. The present invention can effectively overcome the control performance degradation caused by valve nonlinearity, the scheme is simple to implement, and has important practical value in improving the control quality of the automatic control loop.

Description

Technical Field

[0001] The present invention relates to the field of industrial control and precision electromechanical system control, and is particularly applicable to control scenarios with valve nonlinearity problems in control systems. The method is suitable for solving control system oscillation and control performance degradation problems caused by valve nonlinearity in control systems. Background Art

[0002] Valves are one of the most widely used actuators in process industry production. The performance of valves has a great impact on process control, and is closely related to product quality, energy saving and consumption reduction, production safety, etc. Statistical studies have shown that about 30% of control loops in actual industrial processes suffer from the adverse effects of valve nonlinear characteristics to varying degrees. Common valve nonlinear model structures include "dead zone", "variable gain", "hysteresis", "hysteresis loop", "gap", "viscosity", etc. Figure 1 The nonlinearity of the valve in the actual system may be caused by Figure 1 It is composed of one or more nonlinear structures.

[0003] Take valve stiction as an example. This phenomenon is one of the most common and difficult to overcome valve nonlinearity problems. It can easily lead to a decline in the performance of the automatic control system, cause control loop oscillation, and ultimately lead to reduced product quality and equipment damage. When dealing with valve stiction nonlinearity problems, actual industrial control systems mostly use the method of weakening controller action to achieve automatic control. The literature (Mohammad MA, Huang B. Compensation of control valvestiction through controller tuning [J]. Journal of Process Control, 2012, 22 (9): 1800-1819.) introduces how to adjust the PID controller to make the control system stable. This method sacrifices control performance in exchange for system stability and has obvious limitations. Another method is knocker compensation (see TA friction compensator for pneumatic control valves [J]. Journal of Process Control, 2002, 12 (8): 897-904.), by superimposing a jitter square wave signal on the control signal, the stiction problem is overcome by increasing the energy of the corresponding valve input signal. This method can achieve better control performance, but it increases the vibration frequency of the valve during the control compensation process, greatly reducing the service life of the valve, so this type of method is rare in practical applications. In addition to the knocker method, another common compensation method is the two-step method (see Srinivasan R, Rengaswamy R. Approaches for efficient stiction compensation in process control valves [J]. Computers & Chemical Engineering, 2008, 32 (1): 218-229.), which requires the knowledge of the valve's stiction model and the controller's steady-state position.

[0004] In overcoming the nonlinear problem of valves, it is necessary to comprehensively consider the control performance and valve service life, which will have a positive effect on improving the efficiency and saving costs in the industrial production process. Summary of the invention

[0005] The purpose of the present invention is to provide a cascade control method based on valve opening feedback soft measurement to overcome the valve nonlinearity problem and improve the control performance in view of the shortcomings of the prior art.

[0006] Commonly used closed-loop control systems in actual industrial sites include Figure 2 As shown, t represents time; r(t) is the set value; y(t) is the controlled variable, that is, the online output of the sensor, such as temperature, liquid level, flow, pressure, etc.; u(t) is the valve opening instruction; x(t) is the actual valve opening or valve feedback. Note that the actual valve opening x(t) is usually not measured online, or the online measuring instrument is unreliable. For similar industrial control systems, the cascade control scheme based on valve opening feedback soft measurement provided by the present invention can solve a variety of valve nonlinear problems. The proposed method includes the following steps:

[0007] (1) Establish a valve nonlinear model from valve opening command u(t) to valve opening feedback x(t)

[0008] (2) Constructing soft measurement signal for valve opening feedback The sub-loop control system adjusts the parameters of the sub-loop controller to make the control performance meet the requirements;

[0009] (3) Connect the sub-loop to the main loop in the form of cascade control, and adjust the parameters of the main loop controller so that the control performance of the controlled variable meets the requirements.

[0010] Furthermore, the valve opening feedback x(t) is the actual valve opening or other process variables reflecting the actual valve opening, such as flow, temperature, pressure, etc.

[0011] Furthermore, the valve nonlinear model It is a nonlinear model from valve opening command u(t) to valve opening feedback x(t).

[0012] Furthermore, dynamic data modeling or system identification methods are generally used to perform nonlinear modeling of valves. J. Modeling and identification of systems with backlash [J]. Automatica, 2010, 46 (2): 369-374.) proposed a valve clearance model estimation method, and the literature (Choudhury, MAA Shoukat, Nina F. Thornhill, and Sirish L. Shah. "A data-driven model for valve stiction." IFAC Proceedings Volumes 37.1 (2004): 245-250.) studied the estimation of valve viscosity model.

[0013] Furthermore, the valve opening feedback soft measurement signal According to the established valve nonlinear model And the valve opening command u(t) is calculated online in real time.

[0014] Furthermore, the setting value of the secondary loop controller is the valve opening setting value x SP (t), the controlled variable is u(t), and the controlled variable is

[0015] Furthermore, the parameters of the secondary loop controller can be obtained by offline simulation trial and error; SP (t) A triangular wave signal or a sine wave signal can be used as the simulation excitation; the parameters of the sub-loop controller are adjusted so that the controlled variable Able to track x stably, unbiasedly and quickly SP (t).

[0016] Furthermore, the sub-loop linearizes the nonlinear characteristics of the valve by controlling the valve opening feedback soft measurement.

[0017] Furthermore, the sub-loop controller includes: a sub-loop PID controller and a sub-loop fuzzy controller.

[0018] Furthermore, the setting value of the main loop controller is the process value r(t), and the operating variable is the valve opening setting value x SP (t), the controlled variable is y(t).

[0019] Furthermore, in the cascade connection between the main loop and the sub-loop, the output of the sub-loop controller is simultaneously connected to the actual valve opening command u(t), the valve nonlinear model The input is connected.

[0020] Furthermore, the main loop controller generally adopts a PID controller; when the main loop PID parameters are adjusted, the sub-loop and the controlled object can be regarded as an equivalent whole; the parameter adjustment can adopt the system identification and internal model control (IMC) adjustment method, see reference (Rivera, Daniel E., Manfred Morari, and Sigurd Skogestad. "Internal model control: PID controller design." Industrial & engineering chemistry process design and development 25.1 (1986): 252-265.).

[0021] The beneficial effects of the present invention are as follows: the present invention provides a cascade control method based on soft measurement of valve opening feedback, constructs a sub-loop for linearizing valve characteristics through soft measurement of valve opening feedback signal, and forms a cascade control structure with the actual controlled object. The control scheme can effectively overcome the valve nonlinearity problem and improve the control performance. The present invention is simple and convenient to implement and has important practical value. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 Common valve nonlinear model structures, (a) is dead zone, (b) is variable gain, (c) is hysteresis, (d) is hysteresis loop, (e) is clearance, and (f) is viscosity;

[0023] Figure 2 This is the structure diagram of the closed-loop control system before transformation;

[0024] Figure 3 A schematic diagram of a valve nonlinear model provided by an embodiment of the present invention;

[0025] Figure 4 The soft measurement signal based on valve opening feedback in the embodiment of the present invention The structure diagram of the cascade sub-loop control system;

[0026] Figure 5 It is a structural diagram of a cascade control system based on valve opening feedback soft measurement in an embodiment of the present invention;

[0027] Figure 6 Schematic diagram of an equivalent single-loop control system in an embodiment of the present invention;

[0028] Figure 7 It is a schematic diagram comparing the control conditions before and after the transformation using the present invention. DETAILED DESCRIPTION

[0029] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0030] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0031] The present invention provides a cascade control method based on valve opening feedback soft measurement to solve various service valve nonlinear problems. The method comprises the following steps:

[0032] (1) Use data modeling or system identification methods to establish a valve nonlinear model from valve opening command u(t) to valve opening feedback x(t) And calculate the valve opening feedback soft measurement signal like Figure 3 As shown;

[0033] (2) Constructing soft measurement signal for valve opening feedback The auxiliary loop control system adjusts the parameters of the auxiliary loop controller to make the control performance meet the requirements; the connection structure is as follows Figure 4 As shown, the setting value of the secondary loop controller is the valve opening setting value x SP (t), the controlled variable is u(t), and the controlled variable is

[0034] (3) Connect the secondary loop to the main loop in the form of cascade control, and adjust the parameters of the main loop controller so that the control performance of the controlled variable meets the requirements; the connection structure is as follows: Figure 5 As shown, x SP (t) is connected to the output of the main loop controller. The set value of the main loop controller is the process value r(t), and the manipulated variable is the valve opening set value x SP(t), the controlled variable is y(t), in the cascade connection between the main loop and the auxiliary loop, the output of the auxiliary loop controller is simultaneously connected with the actual valve opening command u(t), the valve nonlinear model The input is connected.

[0035] In one embodiment, the secondary loop controller adopts a PID controller, that is, a proportional, integral, and differential controller. The form of the PID controller is:

[0036]

[0037] Among them, k C is the proportionality coefficient, t I is the integration time, t D is the differential coefficient.

[0038] The parameters of the secondary loop controller can be obtained through offline simulation and trial and error; the construction is as follows: Figure 4 The closed-loop simulation control system shown in the figure performs sub-loop PID parameter tuning, where x SP (t) A triangular wave signal or a sine wave signal can be used as the simulation excitation; the parameters of the secondary loop PID controller are adjusted so that the controlled variable Able to track x stably, unbiasedly and quickly SP (t).

[0039] In one embodiment, the main loop controller adopts a PID controller. When the main loop PID parameters are adjusted, the sub-loop and the controlled object can be regarded as an equivalent whole, such as Figure 6 As shown in the figure, the parameter setting can be done by system identification and IMC parameter setting methods. When implementing IMC parameter setting for the main loop PID, the system identification method needs to be used to establish x SP (t) to y(t) first-order or second-order linear model; taking the first-order self-balancing process as an example, the linear model obtained is in the form of:

[0040]

[0041] Among them, s is the Laplace operator, k P is the model gain, τ P is the inertia time constant, θ is the time delay;

[0042] Select an appropriate closed-loop response time λ and calculate the main loop PID parameters according to the following formula:

[0043]

[0044] Among them, k C is the proportionality coefficient, τ I is the integration time, τ D is the differential coefficient.

[0045] The present invention is further described in detail below by taking the nonlinear treatment of the backlash of a cooling water regulating valve in a reactor temperature control loop in a domestic fine chemical enterprise as an example.

[0046] After modeling the loop, the controlled object adopts a first-order pure lag model Unmeasurable interference uses white noise through the first-order inertial link The colored noise obtained after that is controlled by PI, and the parameters are The set value r(t) of the process output signal (water flow) is 60m 3 / h (cubic meters per hour).

[0047] Step 1: Establish a nonlinear clearance model of the cooling water valve. Using the system identification method, according to the literature ( J. Modeling and identification of systems with backlash [J]. Automatica, 2010, 46 (2): 369-374.) is used to establish a backlash valve clearance model from the valve opening command u (t) to the actual valve opening feedback x (t), where the parameter Deadband (DB) = 8.5.

[0048] Step 2: Construct a soft measurement signal to control the valve opening feedback The sub-loop PID control system. The sub-loop controller uses the Backlash valve model with DB = 8.5 to operate u(t) Controlled at the actual valve opening setting value x SP (t). The secondary loop controller uses a PI controller with the following parameters:

[0049] Step 3: Connect the secondary loop to the primary loop in the form of cascade control. SP (t) is connected to the output of the main loop controller, and u(t) is connected to the output of the sub-loop controller. After debugging, the PID control form of the main loop is Achieve the required control performance.

[0050] Figure 7 The control comparison of the controlled variables under the same noise environment is shown, where the solid line is the controlled variable setting value, the dashed line is the original system controlled variable curve, and the dotted line is the controlled variable curve after the modification of the present invention. It can be seen that the present invention can effectively overcome the control performance degradation caused by valve nonlinearity and improve the control performance.

[0051] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A control method to overcome valve nonlinearity, characterized in that: The method includes: Establish a valve nonlinear model from valve opening command u(t) to valve opening feedback x(t) Constructing soft-measurement signal for valve opening feedback The sub-loop control system adjusts the parameters of the sub-loop controller to make the control performance meet the requirements; The sub-loop is connected to the main loop in the form of cascade control. The sub-loop linearizes the nonlinear characteristics of the valve by controlling the valve opening feedback soft measurement, and adjusts the parameters of the main loop controller to make the control performance of the controlled variable meet the requirements.

2. A control method for overcoming valve nonlinearity according to claim 1, characterized in that: The valve nonlinear model It is a nonlinear model from valve opening command u(t) to valve opening feedback x(t).

3. A control method for overcoming valve nonlinearity according to claim 1, characterized in that: The valve opening feedback x(t) is the actual valve opening or a process variable reflecting the actual valve opening.

4. A control method for overcoming valve nonlinearity according to claim 1, characterized in that: The valve opening feedback soft measurement signal According to the established valve nonlinear model And the valve opening command u(t) is calculated online in real time.

5. A control method for overcoming valve nonlinearity according to claim 1, characterized in that: The setting value of the secondary loop controller is the valve opening setting value x SP (t), the controlled variable is u(t), and the controlled variable is 6. A control method for overcoming valve nonlinearity according to claim 5, characterized in that: The secondary loop controller operates u(t) to make Able to track x stably, unbiasedly and quickly SP (t).

7. A control method for overcoming valve nonlinearity according to claim 1, characterized in that: The sub-loop controller includes: a sub-loop PID controller and a sub-loop fuzzy controller.

8. A control method for overcoming valve nonlinearity according to claim 1, characterized in that: The setting value of the main loop controller is the process value r(t), and the operating variable is the valve opening setting value x SP (t), the controlled variable is y(t).

9. A control method for overcoming valve nonlinearity according to claim 1, characterized in that: In the cascade connection between the main loop and the sub-loop, the output of the sub-loop controller is simultaneously connected to the actual valve opening command u(t) and the valve nonlinear model The input is connected.

Images