Skip to Content

Register Login

Industry Online Support

Technical Forum

Navigation

Using continuous PID FB41 for nitrogen pressure regulation

Created by: aag_03 at: 10/5/2016 2:57 PM (8 Replies)

Rating (2)

Thanks 2

9 Entries

10/5/2016 2:57 PM	Rate (0)
aag_03 Posts: 84 Rating: (2)	Hello dear experts! I am a little bit new in continuous controlling with SIMATIC, so I am confused is my code wrong or not. I have a nitrogen pressure sensor (4-20 mA) connected to PLC S7-300 AI and I need stabilize a nitrogen pressure in the tube to some setpoint value. Setpoint is set by operator via HMI. Nitrogen flow in tube is controlled by a valve with analog positioner and position feedback (all 4-20 mA). For valve control I plan to use PLC AO 4-20 mA and AI 4-20 mA for read feedback valve position signal. And I plan to use cont_c block FB41 for PID regulation of pressure. I call FB41 in OB 35 (100 ms). I scale my process value (#pv_val) and setpoint (#sp_val) from 0-2.5 MPa to 0-100% and connect them to appropriate inputs of FB41. Would you please check my further code of FB41 call for rightness? I test it with PLCSIM and it seems to work, but is it work right or not - there is a question? Especially I confuse about CYCLE and GAIN parameters values. Great thanks for any help!!! Piece of code in OB35: CALL "CONT_C" , DB4108 COM_RST := MAN_ON :="K1_8a_Mode_Manual" //valve manual mode PVPER_ON:= "Log_0" //always zero P_SEL := I_SEL := INT_HOLD:= I_ITL_ON:= D_SEL := CYCLE :=T#100MS SP_INT :=#sp_val //pressure setpoint scaled in 0-100% PV_IN :=#pv_val //current pressure value scaled in 0-100% PV_PER := MAN :="from_SCADA_DB".K1_8a_man_setpoint //manual setpoint of valve position in 0-100% GAIN :=-2.000000e+000 // minus as of closing valve for high pressure and opening for low pressure TI := TD := TM_LAG := DEADB_W := LMN_HLM := LMN_LLM := PV_FAC := PV_OFF := LMN_FAC := LMN_OFF := I_ITLVAL:= DISV := LMN :="K1_8a_Reg_Out_prc" //valve setpoint output in % LMN_PER :="AO_DB".POS_1_8a //valve setpoint output for AO QLMN_HLM:= QLMN_LLM:= LMN_P := LMN_I := LMN_D := PV := ER :=
aag_03 Posts: 84 Rating: (2)	Last edited by: aag_03 at: 10/6/2016 12:58:08 PM
Suggestion To thank Quote Answer

10/7/2016 11:35 AM	Rate (0)
CaOd Platinum Member Joined: 5/6/2010 Last visit: 6/9/2023 Posts: 1887 Rating: (255)	Hello aag_03, at the first glance I miss a few values and informations. Do you need a regulator at P, PI or PID? You use the prefabricated values at the instance DB for TI and TD. I would define also the values of TI and TD. Your value at the GAIN is in my opinion to high. You should start with lower than 0.5 and should change the values of GAIN; TI and TD at the comissioning Do you need a fast reaction of the regulator? If not then I would disable the D_SEL, Then you have a PI-regulator. The cycle is ok and must be the same or a multiple value of the calling time of OB35. Your value is absolutly ok. This is only at the first glance. regards CaOd

Suggestion To thank Quote Answer

10/8/2016 11:05 AM	Rate (0)
aag_03 Posts: 84 Rating: (2)	Hello CaOd! Thank you very much for reply! Now this became more clear for me a little :) All values which I use in FB41 are prefabricated. See instance DB of FB41 in an attachment. The type of regulator is PI. The derivative gain is off now, but I plan to use it i future if is necessary and if I will need a faster reaction. As you say I plan to adjust GAIN, TI, TD and other settings values during commissioning. My last questions are about GAIN. Now its value is negative (-2.0). Because I need to close my valve to make pressure higher and open valve to make it lower. Is it right or not? And its prefabricated value is 2.0. Your advice is to start adjusting regulation with GAIN value lower 0.5 (or -0.5 ?). Could you please explain why? Attachment DB4108.pdf (215 Downloads)
aag_03 Posts: 84 Rating: (2)
Suggestion To thank Quote Answer

10/8/2016 8:03 PM	Rate (1)
CaOd Platinum Member Joined: 5/6/2010 Last visit: 6/9/2023 Posts: 1887 Rating: (255)	Hello aag_03, you be right with the - at the GAIN when you must close your valve if you need a higher pressure. Then is my suggestion also -0.5. This value is a experience to start a comissioning with a low risk, somteimes is 0.3 (-0.3) better. If you use a value higher than 1, then is the reaction of the regulator sometimes a little bit difficult. The difference between setpoint and the actual process value go with the factor into the calculation of the output of the regulator (higer GAIN - stronger reaction). The reaction of the regulator could be very strongand it could come to swinging the output signal. If you not need a fast reaction then it's better the value of GAIN is lower than 1 (or -1 at your application). You have than more control about the reaction of the rgulator. The next suggestion is that you limit the output signal of the regulator with the signals LMN_HLM and LMN_LLM. That is also for a better control of the regulator during comissioning and later in the process. You should also use the Deadband (signal DEADB_W) to stabilize your regulator. I hope this is a little bit helpful for you. If you have more questions let me know. regards CaOd

Suggestion To thank Quote Answer
This contribution was helpful to 1 thankful Users aag_03

10/10/2016 7:53 AM	Rate (0)
aag_03 Posts: 84 Rating: (2)	Hello CaOd! I am very appreciative to you for so great and experienced explanation! This is really helpful for me :) But I have once more questions. Is it normal that FB41 regulator does not control feedback valve position? I put target pressure value (scaled 0-100%) on SP input, current pressure value (scaled 0-100%) on PV input, from LMN output I get valve position setpoint 0-100%, but I have not any input for the valve position feedback. How regulator does take it into account?
aag_03 Posts: 84 Rating: (2)	Last edited by: aag_03 at: 10/10/2016 8:37:18 AM
Suggestion To thank Quote Answer

10/11/2016 7:06 AM	Rate (1)
CaOd Platinum Member Joined: 5/6/2010 Last visit: 6/9/2023 Posts: 1887 Rating: (255)	Hello aag_03, the FB41 have no feedback control for the valve. Your target value and your actual value is that what the regulator need for the regulation. With the LMN you limit the output signal for the valve. If the valve have not the right reaction, you can see it at the signal ER. There you can see the actual regulation difference between target and actual value. You should check the difference between this 2 signals and if the difference is to high over a time then you have a error occured. You can also check if the output signal a longer time the same like the LMN, e.g. LMN_HLM or LMN_LLM, then can you also generate a alarm. I hope this help you again. regards CaOd

Suggestion To thank Quote Answer
This contribution was helpful to 1 thankful Users aag_03

10/11/2016 2:40 PM	Rate (0)
aag_03 Posts: 84 Rating: (2)	Hello, dear CaOd! Thank you very much! Your explanation is great and is really helpful for me! Now everything is clear :)
aag_03 Posts: 84 Rating: (2)
Suggestion To thank Quote Answer

3/9/2017 9:48 AM	Rate (0)
Min_Moderator Platinum Member Joined: 9/3/2014 Last visit: 3/13/2019 Posts: 4812 Rating: (123)	New question published by piotrek1988ab is split to a separate thread with the subject Searching for first steps of PID controllers in S7-300. Best regards Min_Moderator

Suggestion To thank Quote Answer

3/9/2017 9:51 AM	Rate (0)
Min_Moderator Platinum Member Joined: 9/3/2014 Last visit: 3/13/2019 Posts: 4812 Rating: (123)	New question published by norimane is split to a separate thread with the subject PID regulation with FB41 for management of a compressor station. Best regards Min_Moderator

Suggestion To thank Quote Answer

|

Share this page:

Share this page on...

Follow us on