Gearless coal mill anti-interference sliding mode control method based on dead-beat prediction

Abstract

The invention belongs to the field of gearless coal mill control, and discloses a gearless coal mill anti-interference sliding mode control method based on dead-beat prediction. The invention provides a gearless coal mill control strategy based on a fast nonsingular terminal sliding mode, an extended state observer and dead-beat prediction so as to improve the anti-interference performance under low-speed variable working conditions. Firstly, aiming at the problem that the control precision of a system on the rotating speed is insufficient due to power supply voltage fluctuation and environment temperature change in the operation process of the gearless coal mill, the invention provides a rapid nonsingular terminal sliding mode control strategy so as to improve the rotating speed response of the gearless coal mill; and secondly, aiming at the problem of load disturbance caused by different coal qualities, an extended state observer is designed and feedforward compensation is carried out on the disturbance, so that the robustness of the system is improved. And finally, aiming at the problem of parameter mismatch caused by heating and tiny vibration when the coal mill runs for a long time, the invention provides deadbeat current prediction control based on sliding mode control. According to the strategy, the dynamic response speed of dead-beat current prediction control is high, and the high robustness of sliding mode control is reserved.

Description

Technical Field

[0001] The invention relates to a gearless coal mill control method, and more particularly to a non-singular fast terminal sliding mode control and deadbeat predictive control method for a gearless coal mill. Background Art

[0002] As a modern grinding equipment, gearless coal mill plays a core role in thermal power generation and other heavy industrial fields. Its efficient operation depends on the powerful performance of its internal drive motor. Therefore, how to improve the anti-disturbance performance of the internal drive motor of the gearless coal mill has always been a huge problem. Nowadays, the asynchronous motor in the coal mill is gradually replaced by the permanent magnet synchronous motor, which is due to the advantages of simple structure, high operating efficiency and power factor, low loss, and good control ability. However, the different quality of coal will cause the motor to be in variable load operation for a long time, resulting in excessive speed overshoot and poor dynamic performance of the system. In addition, since the coal mill often works under high temperature, the motor model parameter mismatch problem will also occur, affecting the robustness of the system current. Therefore, it is of great theoretical significance and application value to study the control method of the gearless coal mill system with high dynamic performance and strong robustness to variable load disturbance and model parameter mismatch.

[0003] Due to the large overshoot and poor system robustness in the face of disturbances in the traditional PI control strategy, it cannot meet the working requirements of the gearless coal mill. Therefore, researchers have proposed various nonlinear control schemes. These include adaptive control, fuzzy control, neural network control, sliding mode control, etc. Among them, sliding mode variable structure control is more suitable for industrial use due to its advantages such as fast response, insensitivity to parameter changes and disturbances, and simple physical implementation. However, there is a problem with traditional sliding mode control: the approach speed and stability of the system cannot be balanced. That is, when the approach speed is increased by adjusting the sliding mode parameters, greater jitter and overshoot will often be generated, and jitter will reduce the tracking accuracy and dynamic performance of the system. At the same time, the disturbance sources faced by the gearless coal mill are diverse and difficult to predict. It is often difficult to meet the requirements of high precision and high robustness by relying solely on traditional control algorithms. At this time, it is necessary to add feedforward compensation on the basis of the controller to improve the anti-interference performance of the system.

[0004] Deadbeat current predictive control has received extensive attention in recent years due to its simple control process, easy implementation, and excellent dynamic performance. However, the deadbeat current prediction method also needs to face the problem of model parameter mismatch or parameter uncertainty. One of the methods is to use disturbance observation or error compensation measures, such as observers, integral compensation, and feedforward compensation. In addition, another method is to improve the accuracy of motor model parameters, such as introducing parameter identification methods. Although the above methods can improve the ability to cope with parameter mismatch, the implementation process is relatively complicated due to the need to construct a disturbance observer or perform complex calculations. Therefore, it is necessary to design a relatively simple control method to achieve the improvement of the robustness of the deadbeat predictive current control parameters. Summary of the invention

[0005] The present invention proposes a composite control strategy for a gearless coal mill to improve the anti-disturbance performance under low-speed and variable working conditions. First, in terms of the speed loop of the motor, the present invention adopts non-singular fast terminal sliding mode control to enhance the system's speed control accuracy. Secondly, in order to reduce the problem of load disturbance caused by different coal quality in the gearless coal mill, the present invention proposes an expanded observer method to observe unknown load disturbances, and feed back the observed values ​​to the speed controller, thereby reducing the jitter of the speed and torque. This composite control strategy enables the system speed to have both a high approach speed and good stability performance. Finally, in terms of the current loop of the motor, the present invention proposes a deadbeat current prediction method based on sliding mode control, which not only overcomes the traditional PI control method's inability to meet the demand for fast response speed, but also solves the limitations of the deadbeat current prediction method. The proposed method has both the fast dynamic response speed of the deadbeat current prediction control and the strong robustness of the sliding mode control, which well solves the parameter mismatch problem caused by heating and slight vibration of the gearless coal mill during long-term operation.

[0006] The technical solution adopted by the present invention to solve the technical problem is as follows:

[0007] The present invention proposes an anti-disturbance sliding mode control method for a gearless coal mill based on deadbeat prediction, and the method comprises the following steps:

[0008] Step 1: The signal detection circuit 3 detects the three-phase stator current i of the permanent magnet synchronous motor PMSM2. a 、i b 、i c The three-phase static current is calculated by the three-phase static to two-phase static module 12, the two-phase static to two-phase rotating module 11 and the motor rotor position angle θ obtained by the speed and rotor position sensor 13 to obtain the d-axis current i d , q-axis current i q .

[0009] Step 2: Get i from step 1q and the speed ω obtained by the rotor position sensor 13 m The disturbance estimation value d is calculated by the extended observer 9. Then the speed ω obtained by the speed and rotor position sensor 13 is m With given speed ω * The difference x, i obtained in step 1 q , and the disturbance estimates d and ω calculated by the extended observer 9 m As the input of the non-singular fast terminal sliding mode speed controller, the output q-axis current As the input of the deadbeat current prediction controller 6 based on sliding mode control.

[0010] Step 3: Set the given The result obtained in step 2 The i obtained by the one-shot delay compensation module 10 d (k+1), i q (k+1) and the speed ω obtained by the rotor position sensor 13 m As input, the d-axis voltage u is calculated by the deadbeat current prediction controller 6 based on sliding mode control. d , q-axis voltage u q The output u d 、u q and i obtained in step 1 d 、i q , and the speed ω obtained by the rotor position sensor 13 m As the input of the one-shot delay compensation module 10.

[0011] Step 4: Substitute u obtained in step 3 d 、u q The voltage component u is calculated by converting the two-phase rotation into a two-phase stationary coordinate system. α 、u β , will u α 、u β As the input of SVPWM module 8, and selecting the corresponding voltage space vector, the control of the permanent magnet synchronous motor non-singular terminal sliding mode, extended observer and zero-beat current prediction is realized.

[0012] The beneficial effects of the present invention are as follows:

[0013] 1) The present invention aims at the problems of large overshoot and slow speed tracking under the complex and changeable working conditions of the gearless coal mill, and therefore proposes sliding mode control to replace the traditional PI control. Singularity is avoided by constructing a non-singular sliding surface, that is, there are no singular points or zero vectors on the sliding surface, ensuring that the control signal is finite and continuous in all states. This method can solve the inherent jitter problem after the system enters the sliding mode stage, accelerate the response speed of the system, and enhance the robustness and stability of the system.

[0014] 2) The present invention aims at the problem of uncertain load disturbance caused by fluctuations in coal quality and other factors in gearless coal mills. The present invention adopts an extended observer to achieve accurate observation of disturbances and reduce the impact of excessive initial estimated peak values ​​on the system. By real-time observation of load disturbances, the observed values ​​are fed back to the speed controller, thereby reducing the jitter of speed and torque.

[0015] 3) The present invention is aimed at the application of PI controller in the current inner loop controller. In the face of the motor model parameter mismatch problem that occurs when the gearless coal mill is often working under high temperature, the present invention proposes a deadbeat current prediction controller based on sliding mode control. The advantage of this controller is that it has the high dynamic response performance of deadbeat current prediction control while retaining the strong robustness of sliding mode control. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Attached Figure 1 It is a structural block diagram of the permanent magnet synchronous motor system of the present invention;

[0017] In the figure: 1. Inverter, 2. Permanent magnet synchronous motor, 3. Signal detection circuit, 4. Control circuit, 5. Non-singular fast terminal sliding mode speed controller, 6. Deadbeat current prediction controller based on sliding mode control, 7. Two-phase rotating to two-phase stationary coordinate system module, 8. SVPWM (voltage space vector pulse width modulation) module, 9. Expanded observer, 10. One-beat delay compensation, 11. Two-phase stationary to two-phase rotating coordinate system module, 12. Three-phase stationary to two-phase stationary coordinate system module, 13. Speed ​​and position sensor, 14. Main circuit. DETAILED DESCRIPTION

[0018] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0019] like Figure 1 As shown, this embodiment includes a main circuit 14, a control circuit 4 and a signal detection circuit 3. The main circuit includes an inverter 1 and a permanent magnet synchronous motor 2; modules 5 to 13 in the control circuit are control algorithms implemented by software. The specific implementation steps are as follows:

[0020] Step 1: The signal detection circuit 3 detects the three-phase stator current i of the permanent magnet synchronous motor PMSM2. a 、i b 、i c The three-phase static current is calculated by the three-phase static to two-phase static module 12, the two-phase static to two-phase rotating module 11 and the motor rotor position angle θ obtained by the speed and rotor position sensor 13 to obtain the d-axis current i d , q-axis current i q .

[0021] The specific algorithm is as follows:

[0022] (1) The algorithm of the three-phase stationary to two-phase stationary coordinate system module 12 is as follows:

[0023]

[0024] (2) The algorithm of the two-phase stationary to two-phase rotating coordinate system module 11 is as follows:

[0025]

[0026] Step 2: Get i from step 1 q and the speed ω obtained by the rotor position sensor 13 m The disturbance estimation value d is calculated by the extended observer 9. Then the speed ω obtained by the speed and rotor position sensor 13 is m With given speed ω * The difference x, i obtained in step 1 q , and the disturbance estimates d and ω calculated by the extended observer 9 m As the input of the non-singular fast terminal sliding mode speed controller, the output q-axis current As the input of the deadbeat current prediction controller 6 based on sliding mode control.

[0027] The specific algorithm is as follows:

[0028] (1) The algorithm of the expanded observer module 9 is as follows:

[0029] Establish the mathematical model of permanent magnet synchronous motor:

[0030]

[0031] Where L d , L q are the d-axis and q-axis stator inductances, respectively, n is the number of motor pole pairs, ω e is the electrical angular velocity of the motor, ω m is the mechanical angular velocity of the motor, T e is the electromagnetic torque of the motor, T Lis the external torque applied by the motor, ψ f is the rotor flux, J is the moment of inertia, and B is the viscosity coefficient.

[0032] Rewrite equation (3) as:

[0033]

[0034] Set the d-axis reference current To obtain the maximum torque-current ratio. This paper takes the surface-mounted permanent magnet synchronous motor as the research object (L d =L q =L), so we can get the following equation:

[0035]

[0036] Assuming that iq is a safe value with a constant a>0, the current that a general motor can withstand is usually selected to be 2 to 3 times the safe value.

[0037] Let: x = ω * -ω m ,ω * is the reference speed.

[0038] According to the mathematical model of permanent magnet synchronous motor, the following control model is derived:

[0039]

[0040] Where d = T L .

[0041] Since there are disturbances in the control system (8), it is necessary to estimate the disturbances in equation (8).

[0042] Design the extended observer (9) to estimate d and It is assumed that the disturbance d(t) is continuous.

[0043]

[0044] fal() is a classic nonlinear function, expressed as follows:

[0045]

[0046] Among them, z2 and They are d and The estimated value of z1 represents the speed estimate, e1=z1-ω e , λ0 and λ1 are the ESO observer filter coefficients.

[0047] (2) The algorithm of the non-singular fast terminal sliding mode speed controller module 5 is as follows:

[0048] Define the following non-singular terminal sliding surface:

[0049]

[0050] The system control law is designed as:

[0051]

[0052] Where α, β, p, q, m, n, η, μ are controller parameters, where μ>0, η>0, they affect the chattering and approach rate of sliding mode control; 1<p / q<2, m / n>1; α, β are sliding mode parameters, which determine the decay rate of trajectory error and the bandwidth of sliding mode dynamics control.

[0053] Step 3: Set the given The result obtained in step 2 The i obtained by the one-shot delay compensation module 10 d (k+1), i q (k+1) and the speed ω obtained by the rotor position sensor 13 m As input, the d-axis voltage u is calculated by the deadbeat current prediction controller 6 based on sliding mode control. d , q-axis voltage u q The output u d 、u q and i obtained in step 1 d 、i q , and the speed ω obtained by the rotor position sensor 13 m As the input of the one-shot delay compensation module 10.

[0054] The specific algorithm is as follows:

[0055] (1) The algorithm of the one-shot delay compensation module 10 is as follows:

[0056]

[0057] Where K pq 、i d (k+1), i q (k+1) represents the one-beat delay compensation current of the dq axis, i d 、i q Indicates the measured current, u d 、u q Indicates the measured voltage.

[0058] (2) The algorithm of the deadbeat current prediction controller 6 based on sliding mode control is as follows:

[0059] According to the voltage equation (3) and the deadbeat current control principle, the voltage reference can be predicted to be

[0060]

[0061] Define the sliding surface as follows:

[0062]

[0063] in Represents the reference current on the dq axis.

[0064] The proposed sliding mode current control function u dq It consists of two parts, namely u smcdq . They satisfy the following expression:

[0065]

[0066] In the designed control function (16), It is the same as the traditional deadbeat current prediction control expression (14). On the other hand, u smcdq Designed to:

[0067]

[0068] Where L is the inductance; ε and k are sliding mode parameters, which determine the decay rate of trajectory error and the sliding mode dynamics control bandwidth.

[0069] Step 4: Substitute u obtained in step 3 d 、u q The voltage component u is calculated by converting the two-phase rotation into a two-phase stationary coordinate system. α 、u β , will u α 、u β As the input of SVPWM module 8, and selecting the corresponding voltage space vector, the control of the permanent magnet synchronous motor non-singular terminal sliding mode, extended observer and zero-beat current prediction is realized.

[0070] The specific algorithm is as follows:

[0071] The specific algorithm of the two-phase rotating to two-phase stationary coordinate system module 7 is as follows:

[0072]

Claims

1. A gearless coal mill anti-disturbance sliding mode control method based on deadbeat prediction, characterized in that: The gearless coal mill control method comprises the following steps: Step 1: The signal detection circuit 3 detects the three-phase stator current i of the permanent magnet synchronous motor PMSM2. a 、i b 、i c The three-phase static current is calculated by the three-phase static to two-phase static module 12, the two-phase static to two-phase rotating module 11 and the motor rotor position angle θ obtained by the speed and rotor position sensor 13 to obtain the d-axis current i d , q-axis current i q . Step 2: Get i from step 1 q and the speed ω obtained by the rotor position sensor 13 m The disturbance estimation value d is calculated by the extended observer 9. Then the speed ω obtained by the speed and rotor position sensor 13 is m With given speed ω * The difference x, i obtained in step 1 q , and the disturbance estimates d and ω calculated by the extended observer 9 m As the input of the non-singular fast terminal sliding mode speed controller, the output q-axis current As the input of the deadbeat current prediction controller 6 based on sliding mode control. Step 3: Set the given The result obtained in step 2 The i obtained by the one-shot delay compensation module 10 d (k+1), i q (k+1) and the speed ω obtained by the rotor position sensor 13 m As input, the d-axis voltage u is calculated by the deadbeat current prediction controller 6 based on sliding mode control. d , q-axis voltage u q The output u d 、u q and i obtained in step 1 d 、i q , and the speed ω obtained by the rotor position sensor 13 m As the input of the one-shot delay compensation module 10. Step 4: Substitute u obtained in step 3 d 、u q The voltage component u is calculated by converting the two-phase rotation into a two-phase stationary coordinate system. α 、u β , will u α 、u β As the input of SVPWM module 8, and selecting the corresponding voltage space vector, the control of the permanent magnet synchronous motor non-singular terminal sliding mode, extended observer and zero-beat current prediction is realized.

2. The permanent magnet synchronous machine control method according to claim 1, characterized in that: The algorithm design of the three-phase stationary to two-phase stationary coordinate system module and the two-phase stationary to two-phase rotating coordinate system module in step 1 includes: (1) The algorithm of the three-phase stationary to two-phase stationary coordinate system module 12 is as follows: (2) The algorithm of the two-phase stationary to two-phase rotating coordinate system module 11 is as follows:

3. The permanent magnet synchronous machine control method according to claim 1, characterized in that: The extended observer design and non-singular fast terminal sliding mode speed controller design in step 2 include: Establish the mathematical model of permanent magnet synchronous motor: Where L d , L q are the d-axis and q-axis stator inductances, respectively, n is the number of motor pole pairs, ω e is the electrical angular velocity of the motor, ω m is the mechanical angular velocity of the motor, T e is the electromagnetic torque of the motor, T L is the external torque applied by the motor, ψ f is the rotor flux, J is the moment of inertia, and B is the viscosity coefficient. Rewrite equation (3) as: Set the d-axis reference current To obtain the maximum torque-current ratio. This paper takes the surface-mounted permanent magnet synchronous motor as the research object (L d =L q =L), so we can get the following equation: Assuming that iq is a safe value with a constant a>0, the current that a general motor can withstand is usually selected to be 2 to 3 times the safe value. Let: x = ω * -ω m ,ω * is the reference speed. According to the mathematical model of permanent magnet synchronous motor, the following control model is derived: Where d = T L . Since there are disturbances in the control system (8), it is necessary to estimate the disturbance in (8). The extended observer (9) is designed to estimate d and It is assumed that the disturbance d(t) is continuous. fal() is a classic nonlinear function, expressed as follows: Among them, z2 and They are d and The estimated value of z1 represents the speed estimate, e1=z1-ω e , λ0 and λ1 are the ESO observer filter coefficients. Define the following non-singular terminal sliding surface: The system control law is designed as: Where α, β, p, q, m, n, η, μ are controller parameters, where μ>0, η>0, they affect the chattering and approach rate of sliding mode control; 1<p / q<2, m / n>1; α, β are sliding mode parameters, which determine the decay rate of trajectory error and the bandwidth of sliding mode dynamics control.

4. The permanent magnet synchronous machine control method according to claim 1, characterized in that: The one-beat delay compensation design and the deadbeat current prediction controller based on sliding mode control in step 3 include: Where K pq 、i d (k+1), i q (k+1) represents the one-beat delay compensation current of the dq axis, i d 、i q Indicates the measured current, u d 、u q Indicates the measured voltage. According to the voltage equation (3) and the deadbeat current control principle, the voltage reference can be predicted to be Define the sliding surface as follows: in Represents the reference current on the dq axis. The proposed sliding mode current control function u dq It consists of two parts, namely u smcdq . They satisfy the following expression: In the designed control function (16), It is the same as the traditional deadbeat current prediction control expression (14). On the other hand, u smcdq Designed to: Where L is the inductance; ε and k are sliding mode parameters, which determine the decay rate of trajectory error and the sliding mode dynamics control bandwidth.

5. The permanent magnet synchronous machine control method according to claim 1, characterized in that: The design of the two-phase rotating to two-phase stationary coordinate system in step 4 includes:

Images