CN113385006A

CN113385006A - Automatic control method for flue gas denitration

Info

Publication number: CN113385006A
Application number: CN202011403564.1A
Authority: CN
Inventors: 陈亮; 李�杰; 刘峥嵘; 冯浩; 闫伟; 翟庆超; 张志宇; 李维虎
Original assignee: Zhongtai Power Plant Of Huaneng Shandong Power Generation Co ltd; Huaneng Jining Canal Generating Co ltd; Huaneng Shandong Power Generation Co Ltd
Current assignee: Zhongtai Power Plant Of Huaneng Shandong Power Generation Co ltd; Huaneng Jining Canal Generating Co ltd; Huaneng Shandong Power Generation Co Ltd
Priority date: 2020-12-04
Filing date: 2020-12-04
Publication date: 2021-09-14

Abstract

A three-way crack vibration frequency measurement and check method is characterized in that an actuator is used for driving an ammonia injection adjusting valve to control ammonia injection amount, ammonia and air are mixed and then injected into an SCR (selective catalytic reduction) to fully react with NOx, a CEMS (continuous emission measurement system) analysis device detects real-time emission of NOx, and then the opening instruction of the actuator of the ammonia injection adjusting valve is determined through calculation. The dynamic optimal IMC-PID control system developed and applied by the project has an automatic optimization function and automatic and manual setting functions, is excellent in control effect, can meet the requirement of continuous automatic control of the on-site NOX, can be realized in a SAMA diagram platform of a DCS, is very convenient to configure and set, does not generate additional equipment cost and maintenance cost, and has good economic applicability.

Description

Automatic control method for flue gas denitration

Technical Field

The invention relates to the field of automatic control of thermal power generation, in particular to an automatic control method for flue gas denitration.

Background

In the SCR flue gas denitration process of a thermal power generating unit, the most important control belongs to ammonia injection control. If the ammonia injection amount is too low, the NOx content at the outlet of the SCR is relatively increased, the denitration efficiency is reduced, and the emission reduction requirement cannot be met; if the ammonia injection amount is too high, the ammonia escape amount will be increased, which is not only uneconomical, but also causes the blockage of the air preheater and the secondary pollution of the emission of toxic ammonia due to the crystallization generated by the reaction of the low-temperature region after the SCR reaction region and the acidic substances in the flue gas. Therefore, the control of NOX in a reasonable range is a necessary prerequisite for safe, environment-friendly and economic operation of the unit. At present, when a power generation plant unit operates in a continuous variable load and AGC-R mode, various traditional control system algorithms are difficult to complete denitration automatic control with high quality due to strong disturbance in various aspects.

Disclosure of Invention

The invention aims to overcome the defects of the prior art, and provides the automatic control method for flue gas denitration, which is stable, reliable, safe, efficient, high in universality and low in maintenance amount, can obviously improve the automatic control quality of NOX, prevents NOX at the total flue gas outlet and excessive safe escape discharge amount, and can further prolong the service life of air preheater equipment.

The method comprises the following steps:

an automatic control method for flue gas denitration is characterized in that in a control system, important parameters capable of approximately reflecting the law of a field NOx generation process are extracted by constructing and utilizing the dynamic relation between the opening of an ammonia injection adjusting door and the NOx emission, and the important parameters are converted into P, I, D process parameters in a PID control structure through an operation method; and meanwhile, the dynamic sensitivity of the PID control system is adjusted in real time by utilizing the dynamic deviation relation between the NOX content detected by a CEMS instrument and an ideal given value, data output by the control system is sent to an actuator control instruction operation unit, an opening instruction of an ammonia injection actuator is obtained through continuous calculation, and then a flue gas denitration control SAMA diagram of the denitration control system is constructed to realize field application and perfect undisturbed switching and following functions.

Further, the control system is designed as a mathematical model control system of PID structure. In the control flow chart, the input of the control process of the NOx is expressed as the real-time NOx monitoring amount AT, when the operator gives the control target amount SP of the NOx, the control process outputs and expresses as a denitration actuator ammonia injection regulating valve instruction C through a PID control algorithm, after the denitration actuator instruction opening C is established, an ammonia injection regulating valve is opened, ammonia-air mixed gas is sent to the SCR and reacts with the NOx, the NOx emission amount AT is changed to approach the NOx and finally stabilize the NOx near the SP, and the industrial process is the nitrogen oxide control process. The detection of the NOx emission adopts an in-situ nano semiconductor NOx monitoring system (CEMS). Performing system identification on the nitrogen oxide control process: solving the transfer rule between the ammonia injection amount C and the NOx monitoring amount AT to obtain a mathematical model MI of the theoretical discharge amount of NOx, and performing system identification by adopting a disturbance test method to obtain the transfer rule of the control process of the nitrogen oxide, wherein the transfer rule comprises the following steps: increasing gain, namely changing C by x%, obtaining that AT is changed by y%, and then changing the gain KI to y/x; the delay time T, namely the time T2-T1 elapses in the middle from the start of the step disturbance time T1 of the actuator C until the NOx emission AT changes from the time T2, the inertia time T, namely the AT changes from the time T2 to the end of the time T3, the process time elapsed in the middle, the process curve of the NOx is regarded as the second-order inertia characteristic through disturbance tests AT the rated and steady state conditions, the total inertia time range is 150-200 seconds, and the delay time is delayedThe range of the time is 80-120 seconds, the range of the transfer gain is 2-8, the total inertia time value is 480 seconds, the delay time value is 100 seconds, the transfer gain value is 5, and for the second-order inertia process, the inertia time of each-order inertia link is 160 seconds, so the transfer function of the mathematical model MI of the nitrogen oxide control process is as follows:

and processing the inertia time T, the delay time T and the transfer gain K parameters to obtain P, I, D parameters of the PID control algorithm. The specific treatment process is as follows:

proportional gain P of PID is 2 × T ÷ X;

integral gain I of PID is 1 ÷ X;

the differential gain D of PID is T squared ÷ X;

in the above formula, X ═ K (T + y × T);

Y＝(Z*0.45+1.5)÷(Z*1.45-1.5).

further, Z is a real number, varying in real time within the range of 1.2-2.0.

The basis of the change is as follows: the deviation amount AP-SP of the NOx real-time monitoring amount AT and the NOx control target amount SP meets the following table:

when AP > SP, the 8 second deviation change rate AP-SP satisfies the following table:

(AP-SP) -8 sec ago (AP-SP)	Output of
		-0.200000	0.500000
-0.150000	0.500000
		-0.100000	0.600000
-0.050000	0.700000
		0.000000	0.750000
0.050000	0.800000
		0.100000	0.900000
0.150000	1.000000
		0.200000	1.000000

When AP < SP, the 8 second deviation change rate AP-SP satisfies the following table:

the amount of change and the rate of change of the AP and the SP are multiplied to obtain data in a range of (0.25-1), and the data is processed as follows to obtain the parameter Z.

Furthermore, the applied DCS platform is an LN2000 system, and various basic operation functional blocks and PID functional blocks in the LN2000 system are used for calculating the actuator instruction opening degree C.

The invention has the beneficial effects that:

the method has high universality and can be directly implemented in DCS platforms of various brands/models.

2, the automatic denitration control process can be continuously, efficiently and safely finished with high quality.

3 the maintenance amount is extremely low.

4, the ammonia injection can be further saved, namely, the denitration agent is saved.

And 5, the service life of the air preheater equipment is further prolonged.

6, the NOx of the total smoke outlet and the safety escape discharge amount are prevented from exceeding the standard for a long time,

7, the pressure of operators can be effectively reduced, and the productivity is improved.

The implementation, functional features and advantages of the objects of the present invention will be further explained with reference to the accompanying drawings.

Drawings

FIG. 1 is a schematic diagram of the overall plant operation in an automatically controlled process for denitration of flue gases;

FIG. 2 is a structural diagram of an internal model control graph equivalent-to-feedback control in an automatic control method for flue gas denitration;

fig. 3 is a schematic diagram of a simulation result of adjusting a second-order lag object by using dynamic optimal IMC-PID in an automatic control method for flue gas denitration.

FIG. 4 is a schematic diagram of disturbance characteristics when Ms is selected as dynamic self-alignment in an automatic control method for flue gas denitration.

Fig. 5 is a schematic diagram of a dynamic equilibrium process after a large disturbance is given in an automatic control method for flue gas denitration.

FIG. 6 is a schematic diagram of the regulation trend of the field dynamic optimal IMC-PID in the automatic control method for flue gas denitration.

FIG. 7 is a schematic diagram of the regulation trend of the field dynamic optimal IMC-PID in the automatic control method for flue gas denitration.

Detailed Description

1. The defects of the prior art are explained in detail:

the ID controller is a mature control algorithm which is widely applied to the field in the current power generation industry, but has a plurality of problems of large delay, large inertia, poor control dynamic quality of a time-varying system and the like. Therefore, automatic control of NOx in industrial sites is not easy for long-term high-quality continuous investment. The concrete conditions are as follows:

1. since the reaction of NH3 with NOx is a large-delay, large-inertia process, in such a large-delay system, the NOx monitoring amount always needs to be changed after a long delay time after the ammonia injection flow rate regulating valve is changed.

The filtering-continuous sampling-condensing-secondary filtering-chemical analysis process of the CEMS also has obvious delay, and the detected NOx value has unexpected trend in the CEMS purging process, so that the controlled NOx object also has certain hysteresis and nonlinear characteristics.

3. The change of the NOx content of the boiler exhaust smoke is a very complicated process, is greatly influenced by parameters such as unit load, air quantity, coal type, SOFA air door opening degree of the top of the low-nitrogen combustor and the like, and has different delay characteristics; in an AGC peak regulation unit, the internal and external disturbance of the system is very large, and the problems of industrial control are difficult to solve by the traditional field control means such as manual control, PID automatic control and the like.

In order to solve the problem that PID is difficult to adapt to a denitration control system, Smith is adopted to estimate the advanced balance large delay characteristic of the PID once, but the control characteristic is very poor when the model is mismatched, and the time-varying characteristic of the denitration control system cannot be responded; therefore, a denitration internal model follow-up control system (IMC) is further researched subsequently, and can still be stably controlled when a system model is mismatched, so that the automatic control of the NOX has good steady-state characteristics and advanced control capability; however, this system also has some problems: the PID controller is not involved, the advance characteristic of a filter balance inverse model needs to be artificially increased, but the selection of filtering has no reference basis, so that the system is difficult to balance in stability and time, and the process of controlling the setting of the system is long and complicated in order to find accurate and quick parameters; meanwhile, the system can not perform feedforward compensation, and only a follow-up mode is needed to achieve the effect of lead balance, but the lead characteristic of follow-up control lags behind the feedforward, so that quick and accurate lead response cannot be achieved.

2. Overview of embodiments:

Further, the control system is designed as a mathematical model control system of PID structure. In the control flow chart, the input of the NOx control process is expressed as the real-time NOx monitoring amount AT when the operator gives NOAnd after the control target quantity SP of the X is obtained, outputting an instruction C expressed as an ammonia spraying adjusting door of a denitration actuator through a PID control algorithm, after the instruction opening C of the denitration actuator is established, opening the ammonia spraying adjusting door, sending ammonia-air mixed gas into an SCR (selective catalytic reduction), reacting with NOx, and changing the NOx emission AT to enable the NOx emission AT to approach and finally stabilize near the SP, wherein the industrial process is a nitrogen oxide control process. The detection of the NOx emission adopts an in-situ nano semiconductor NOx monitoring system (CEMS). Performing system identification on the nitrogen oxide control process: solving the transfer rule between the ammonia injection amount C and the NOx monitoring amount AT to obtain a mathematical model MI of the theoretical discharge amount of NOx, and performing system identification by adopting a disturbance test method to obtain the transfer rule of the control process of the nitrogen oxide, wherein the transfer rule comprises the following steps: increasing gain, namely changing C by x%, obtaining that AT is changed by y%, and then changing the gain KI to y/x; the delay time T, i.e., the time T2-T1 elapses in the middle of the time from the start of the step disturbance time T1 of the actuator C to the start of the change of the NOx emission AT from the time T2, the inertia time T, i.e., the time AT which the AT changes from the time T2, ends AT the time T3, the intermediate elapsed process time, in rated working condition and steady state, through disturbance test, the process curve of NOx is regarded as second-order inertia characteristic, and the total inertia time range is 150 plus 200 seconds, the delay time range is 80-120 seconds, the transfer gain range is 2-8, the total inertia time value is 480 seconds, the delay time value is 100 seconds, the transfer gain value is 5, for the second-order inertia process, the inertia time of each order of inertia link is 160 seconds, so the transfer function of the mathematical model MI of the nitrogen oxide control process is as follows:

proportional gain P of PID is 2 × T ÷ X;

integral gain I of PID is 1 ÷ X;

the differential gain D of PID is T squared ÷ X;

in the above formula, X ═ K (T + y × T);

Y＝(Z*0.45+1.5)÷(Z*1.45-1.5).

further, Z is a real number, varying in real time within the range of 1.2-2.0.

AP-SP	output of
		-100.000000	1.000000
-7.000000	1.000000
		-4.000000	0.750000
-2.000000	0.500000
		0.000000	0.500000
2.000000	0.500000
		4.000000	0.750000
7.000000	1.000000
		100.000000	1.000000

(AP-SP) -8 sec ago (AP-SP)	Output of
		-0.200000	1.000000
-0.150000	1.000000
		-0.100000	0.900000
-0.050000	0.800000
		0.000000	0.750000
0.050000	0.700000
		0.100000	0.600000
0.150000	0.500000
		0.200000	0.500000

3. The specific implementation and operation of the combined drawing are as follows:

the control algorithm of the automatic control system for the dynamic optimal internal model PID denitration developed by the project is obtained based on the evolution and the upgrade of a PID control algorithm and an internal model control algorithm.

Referring to fig. 1, Internal Model Control (IMC) is a novel control strategy based on a process mathematical model, and is currently popularized and applied in the field of denitration in the power generation industry, and the effect is good.

PV is the actual emission amount of clean flue gas NOX, SP is the set value of clean flue gas NOX, Gp is the actual process of on-site chemical reaction, Gm is a mathematical model representing the on-site reaction process (the transmission gain is K, the inertia time constant is T, and the delay time constant is tau), Gcgf is an internal model controller of the on-site process, and Gf is a filter (the time constant is lambda). Gq is the reciprocal of the pure inertia fraction of Gm.

Referring to fig. 2, the internal model control is feedforward control and PID is feedback control. In order to make the internal model control compatible with the PID, the internal model control graph is equivalently converted into a feedback control structure as shown in FIG. 2;

gc in the figure is the overall feedback controller after the equivalent transformation. The denitration controlled object can be regarded as a high-order transfer function process, and the model adaptation can be determined by the IMC zero steady-state deviation characteristic and the static deviation can be eliminated, so that the mathematical model can be simplified into a second-order delay process. According to a theoretical derivation, the transfer function of Gc can be written as follows:

where ξ is the damping coefficient of the second order process. This structure happens to be a transfer function form of the PID controller.

The actual controlled system is changed in real time, and the lower the control system sensitivity (Ms), the more the control quality results can be unaffected in the time varying of the controlled system. In the design of the control system, a theoretical observation point which enables a zero point of a PID transfer function to be equal to a pole of a transfer function of a controlled object is introduced to obtain the optimal control quality. The relationship between this theoretical λ and the preferred sensitivity Ms can be derived from the following empirical formula:

among them, the classical limitation range of the sensitivity Ms is preferably 1.2 to 2.0.

The simulation of the dynamic optimal IMC-PID internal model controller is carried out, and a GE OC6000E DCS system is adopted as a platform. And adjusting a second-order delay object by utilizing the dynamic optimal IMC-PID. The simulation results are shown in fig. 3.

To simulate a large-scale lag object, the second-order inertia time of the controlled process is set to 3s, and the lag time is set to 10 s. The left and right perturbation tests of the upper graph respectively show the dynamic characteristics that Ms is 2.0\1.6\1.4\ 1.2. It is seen from the figure that the dynamic characteristics of the control system are good no matter how much Ms is equal, wherein when Ms is 2.0, the PV adjustment speed is fast, but oscillation attenuation phenomenon exists; as Ms is reduced, PV oscillation is weaker and weaker, and meanwhile, the adjusting time is prolonged; when Ms is 1.2, the control system adjusts very slowly, but no longer decays. This yields:

(1) the IMC-PID inherits the advantages of IMC and has good large delay resistance;

(2) p, I, D three parameters of IMC-PID can automatically calculate the optimal solution, the PID optimal parameter result can be called for observation, and the feed-forward function of the PID control system can be fully utilized for feed-forward regulation. The IMC-PID thus inherits all the advantages of PID control.

The Ms setting value determines the quality control characteristics. In order to further obtain a better control effect, the Ms dynamic self-optimization function is designed, and the Ms is automatically adjusted by monitoring the variation and the variation rate of the adjusted quantity deviation in real time, so that the optimal sensitivity is automatically and dynamically obtained.

In order to verify the effect of the Ms dynamic self-optimization, an extremely large-delay object (inertia time 3s, delay time 20s) is selected for testing, and when the Ms static selection is 1.8, the disturbance characteristic is shown on the left side of the graph 4; when Ms is chosen to be dynamically self-adjusting (maximum Ms varies dynamically between 1.2-1.84), the perturbation behavior is shown on the right side of fig. 4. It can be seen that the difference between the early-stage dynamic characteristic of the selected Ms dynamic self-tuning and the adjusting speed of the static Ms is not large, but the subsequent steady-state quality is greatly superior to the adjusting quality of the static Ms.

The dynamic balancing process after a given large disturbance is shown on the right side of fig. 5. It can be seen that Ms dynamically changes with the trend of PV changes: the initial state Ms is increased, so that the actuator is adjusted as soon as possible, the PV approaches to the SP as soon as possible, and an excellent dynamic process is embodied; when PV gradually reaches the maximum deviation, Ms is sharply reduced, so that the subsequent steady-state process is better. The dynamically optimal IMC-PID exhibits good regulation performance during dynamic balancing of large disturbances.

In conclusion, the IMC-PID regulation process of the dynamic Ms is selected, the regulation rate and the regulation precision are both considered, the characteristic of strong adaptability of the IMC-PID is kept, and the defect of compromise selection of the adaptability and the regulation speed is overcome. Moreover, the value change of the Ms does not change the result that the PV tends to be stable and converged, so that the dynamic optimization process of the Ms in a reasonable range is safe, and the method has the advantages that other control algorithms do not have.

In the field application case:

the control method is practically applied to a No. 5 unit of Huaneng Jining canal Power Generation Co. In order to improve the control quality, an SCR inlet and outlet NOx analysis instrument selects an in-situ nano semiconductor NOx online monitoring device as a regulated quantity PV and a feed-forward quantity.

According to the design of a simulation model, the dynamic optimal IMC-PID control system is realized in the No. 5 unit LN2000 DCS control system, and is selectively switched with the PID and IMC controllers which are designed and put into use in advance.

As shown in fig. 7, the control system was able to make stable adjustments to eliminate static bias when a setpoint step disturbance of 5mg/Nm3 was made.

When the unit load varies from 165MW to 330MW, the NOx bias can still be stably maintained within + -4 mg/Nm3 using a dynamically optimized IMC-PID control system.

It can be seen that the dynamic optimal IMC-PID denitration system fully inherits the advantages of the IMC and the PID, simultaneously, the sensitivity Ms of the system is optimized in real time, the actuator accurately acts in advance according to the internal model device, and the inlet NOX feedforward can be tracked to perform advanced adjustment, so that the dynamic characteristic is optimal on the premise that the NOX trend of the SCR outlet is changed on site in time due to various disturbances.

The project has the advantages that:

the dynamic optimal IMC-PID control system developed and applied by the project has an automatic optimization function and automatic and manual setting functions, is excellent in control effect, can meet the requirement of continuous automatic control of the on-site NOX, can be realized in a SAMA diagram platform of a DCS, is very convenient to configure and set, does not generate additional equipment cost and maintenance cost, and has good economic applicability.

While there have been shown and described what are at present considered the fundamental principles and essential features of the invention and its advantages, it will be apparent to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, but is capable of other specific forms without departing from the spirit or essential characteristics thereof. The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference signs in the claims shall not be construed as limiting the claim concerned.

Furthermore, it should be understood that although the present description refers to embodiments, not every embodiment may contain only a single embodiment, and such description is for clarity only, and those skilled in the art should make the description as a whole, and the embodiments may be appropriately combined to form other embodiments understood by those skilled in the art.

Claims

1. A three-way crack vibration frequency measurement and check method, drive the ammonia injection regulating gate with the actuator and control the ammonia injection amount, ammonia mixes with air and then injects SCR and NOx to react fully, CEMS analytical equipment detects the real-time emission amount of NOx, and then confirm the opening order of the ammonia injection regulating gate actuator through calculating, characterized by that, in the control system, construct and utilize the dynamic relation of ammonia injection regulating gate opening and NOx emission amount, extract the important parameter that can roughly reflect the producing process law of on-the-spot NOx among them, and convert the above-mentioned important parameter into P, I, D process parameter in PID control structure through the operation method; and meanwhile, the dynamic sensitivity of the PID control system is adjusted in real time by utilizing the dynamic deviation relation between the NOX content detected by a CEMS instrument and an ideal given value, data output by the control system is sent to an actuator control instruction operation unit, an opening instruction of an ammonia injection actuator is obtained through continuous calculation, and then a flue gas denitration control SAMA diagram of the denitration control system is constructed to realize field application and perfect undisturbed switching and following functions.

2. The three-way crack vibration frequency measurement and check method according to claim 1, characterized in that: controlThe system is designed as a mathematical model control system of PID structure. In a control flow chart, the input of the control process of the NOx is expressed as the real-time NOx monitoring amount AT, when the operator gives a control target amount SP of the NOx, the control process outputs an instruction C expressed as a denitration actuator ammonia injection regulating valve through a PID control algorithm, after the instruction opening C of the denitration actuator is established, the ammonia injection regulating valve is opened, ammonia-air mixed gas is sent to the SCR and reacts with the NOx, the NOx emission amount AT is changed to approach the NOx and finally stabilize the NOx in the vicinity of the SP, and the industrial process is the nitrogen oxide control process. The detection of the NOx emission adopts an in-situ nano semiconductor NOx monitoring system (CEMS). Carrying out system identification on the nitrogen oxide control process: solving the transfer rule between the ammonia injection amount C and the NOx monitoring amount AT to obtain a mathematical model MI of the theoretical discharge amount of NOx, and performing system identification by adopting a disturbance test method to obtain the transfer rule of the control process of the nitrogen oxide, wherein the transfer rule comprises the following steps: transferring the gain, namely changing C by x%, obtaining that AT is changed by y%, and then changing the gain KI to y/x; the delay time T, namely the actuator C starts step disturbance time T1, the time T is T2-T1 is elapsed from the time when the NOx emission AT starts to change from the time T2, the inertia time T, namely the AT starts to change from the time T2, the time T3 is ended, the process time elapsed from the middle, when the rated working condition is in a steady state, the process curve of the NOx is regarded as second-order inertia characteristic through disturbance test, the total inertia time range is 150-200 seconds, the delay time range is 80-120 seconds, the range of the transfer gain is 2-8, the total inertia time is 480 seconds, the delay time is 100 seconds, the transfer gain value is 5, for the second-order inertia process, the inertia time of each order of inertia links is 160 seconds, therefore, the transfer function of the mathematical model MI of the nitrogen oxide control process is:

proportional gain P of PID is 2 × T ÷ X;

integral gain I of PID is 1 ÷ X;

the differential gain D of PID is T squared ÷ X;

in the above formula, X ═ K (T + y × T);

Y＝(Z*0.45+1.5)÷(Z*1.45-1.5)。

3. the three-way crack vibration frequency measurement and check method according to claim 2, characterized in that: z is a real number and varies in real time within the range of 1.2-2.0;

Multiplying the variation and the variation rate of the AP and the SP to obtain data in a range of (0.25-1), and performing the following data processing to obtain a parameter Z;

。

4. the three-way crack vibration frequency measurement and check method according to claim 1, characterized in that: the applied DCS platform is an LN2000 system, and various basic operation functional blocks and PID functional blocks in the LN2000 system are used for calculating the actuator instruction opening degree C.