Feedforward feedback control system design (process control course design matlab/simulink)

WeChat official account ： Chuangxiang diary
Send keywords ： Feedforward and feedback
Get the full report for free +matlab/simulink Simulation source file

---

One 、 Course design task analysis

The transfer function of a given process is G_PC (s)=(s+1)/(s+2)(2s+3) ,G_PD (s)=5/(s+2) , Controller use PID law . Try to design a feedforward - Mathematical model of feedforward regulator in feedback control system G_ff, And use SIMULINK and MATLAB Program simulation implementation . When the disturbance signal F(s) Unit step signal and SINS The signal , System setpoint R(s) by 1 when , adjustment PID Parameters , It can overcome interference F Impact on the system , It can also track the set value R The change of .（ Suppose the transfer function of the detection transmission link is 1）.

1、 The design requirements
① The model expression of feedforward controller is given ;
② Provide system simulation diagram ;
③ Give the debugging steps of the system after the design is completed , Explanation of experimental results and analysis of experimental data ;
④ If the control channel has large pure delay , namely G_PC (s)=(s+1)/(s+2)(2s+3) e^(-10s), Control how performance changes ？ The improved feedforward - Structure model control scheme and simulation results of feedback hole control system .
⑤ Write the experiment report .

2、 Process analysis
In the system, there is a feedforward control to compensate the main disturbance signal , There is also the problem of using feedback control to overcome other interference signals , Such a system is feedforward - Feedback control system . The function of the feedforward control system is to compensate the main interference signals , It can be aimed at the main interference signal , Set the corresponding feedforward controller . The purpose of introducing feedback control is that the system can overcome the influence of all interference signals on the regulated quantity , In addition to known interference signals , There are other interference signals in the system , These disturbance signals have little influence on the system , Some can be considered , Some cannot be considered or measured , Through feedback control to overcome . The signals to be measured in the system include both modulated and disturbed signals .

---

Two 、 Object property analysis

From the title we can see , The transfer function of the control channel of the system is G_PC (s)=(1/6(s+1))/(1/2 s+1)(2/3 s+1) , The transfer function of the disturbance channel is G_PD (s)=5/(s+2). Therefore, the controlled process is a self balancing multi capacity process , That is, when the input changes , No additional control is required , The process can spontaneously approach a new equilibrium state .

Static gain of the controlled process （ Amplification factor ）K=1/6, That is, the ratio of output variation to input variation , It is easy to know that the amplification factor is small , The control function of the process is weak . According to its time constant T_1=1/2,T_2=2/3, The control process is slow .

Interfere with channel transmission G_PD (s)=5/(s+2), Its static gain K=5/2, It can be seen that the amplification coefficient of the disturbance channel is large , That is, a small disturbance will have a greater impact on the system , Feedforward shall be used - Feedback control to eliminate the influence of disturbance .

---

3、 ... and 、 Control scheme design and simulation

1、 Initial structure of the system
According to the design requirements , The transfer function of the control channel is G_pc(s)=(s+1)/((s+2)(s+3)) And the transfer function of the interference channel is G_pd(s)=5/(s+2), Input set to unit step signal , Do you want to join us PID Controller and feedforward controller , Form a simple single closed-loop system , The original structure block diagram of the system is established as shown in Figure 4-1 Shown .

When the interference input is set to the unit step response signal , The simulation time is set to 20s, In the 10s Add interference when , The system simulation waveform is shown in the figure 4-2 Shown .

From the figure 4-2 You know , When there is no controller , Output cannot follow a given , There is a serious margin , After the interference is added , The anti-interference performance of the system is very poor .
When the interference input is set to a sinusoidal signal , The simulation time is set to 10s, The system simulation waveform is shown in the figure 4-3 Shown .

From the figure 4-3 You know , The system has poor rapidity , And it can not produce stable oscillation after the introduction of sinusoidal signal .

2、 Join in PID Behind the controller
Add... To the system PID The controller keeps the control variable at the set value . Get the system simulation structure diagram 4-4 Shown .

When the interference input is set to the unit step response signal , The simulation time is set to 10s, In the 5s Add interference when , The system simulation waveform is shown in the figure 4-5 Shown .

From the figure 4-5 You know , The system can be fast and stable , However, the stability of the system decreases obviously after the disturbance occurs , There is a large overshoot , And it will take longer to regain stability again .
When the interference input is set to a sinusoidal signal , The simulation time is set to 20s, In the 5s Add interference when , The system simulation waveform is shown in the figure 4-6 Shown .

From the figure 4-6 You know , When the disturbance occurs, the system is affected to produce overshoot and the feedback control function has a certain lag , Disturbance cannot be eliminated in time .

3、 After adding the feedforward controller
A feedforward controller is introduced to measure the disturbance in the process and change the control quantity according to the measured value . Not joined at this time PID controller , Only the feedforward control system is composed as shown in the figure 4-7 Shown .

When the interference input is set to the unit step response signal , The simulation time is set to 20s, In the 5s Add interference when , The system simulation waveform is shown in the figure 4-8 Shown .

From the figure 4-8 You know , The system is little affected by interference when interference occurs , But the system has a large margin , Can't follow the given well .
When the interference input is set to a sinusoidal signal , The simulation time is set to 20s, In the 5s Add interference when , The system simulation waveform is shown in the figure 4-9 Shown .

From the figure 4-9 You know , The system can still ensure the stability of the system in case of sinusoidal signal interference , Make the system suffer little interference , But the system has a large margin , Can't follow the given well .

4、 feedforward - Feedback control system
（ See report for details ）

5、 The system has a large lag
（ See report for details ）

6、 Smith predictor compensator
（ See report for details ）

---

Four 、 Design summary and experience

Through this course design , The three members of our group are responsible for feedforward and feedback control 、PID Parameter setting 、 We have a deeper understanding of the hazards of lagging links and Smith control , Realized the importance of division of labor and cooperation in the group , And apply what we have learned in the textbook to practice . In the process , We also have a lot of problems , I can't adjust to the ideal one many times PID Controller parameters , By searching for information , got it Simulink Can be done PID Self tuning . Understand this , We have abandoned the traditional method of setting parameters , Use Simulink Library Browser Medium Continuous In the module PID controller Self contained Tuner Function to complete PID Automatic parameter setting . Let's talk about MATLAB Have a deeper understanding of the use .

Even though PID The problem of parameter setting has been solved , But at the beginning, the output waveform result is not particularly ideal . In the process , The three members of our group carefully discussed and analyzed the reason through curve comparison , Check every possible problem carefully , Let each person do his best , Finally, a good control effect has been achieved . This process , We have not only learned new knowledge , Consolidate what you have learned , It has also exercised our thinking ability and teamwork ability , This may be the charm of curriculum design .