Multi-fuel combustion boiler SCR (Selective Catalytic Reduction) ammonia injection fuzzy PID (Proportion Integration Differentiation) parallel superposition hybrid control system and method

Abstract

The invention discloses a multi-fuel combustion boiler SCR (Selective Catalytic Reduction) ammonia injection fuzzy PID (Proportion Integration Differentiation) parallel superposition mixing control system and method, and belongs to the field of boiler flue gas pollutant control. The system adopts a hybrid control architecture in which a fixed parameter PID controller and a fuzzy logic controller (FLC) work in parallel. The system calculates the deviation and the change rate of the NOx concentration of an outlet in real time, on one hand, the fixed parameter PID controller outputs the reference ammonia spraying amount, and accurate control under the stable working condition is achieved; on the other hand, the fuzzy logic controller outputs an intelligent compensation control quantity according to the deviation and the change rate, and the intelligent compensation control quantity is specially used for coping with severe working condition changes. And finally, the two outputs are superposed as a total ammonia spraying instruction. According to the method, the adaptive capacity and the control robustness to multi-fuel complex working conditions are improved. The fuzzy controller can perform coarse adjustment quickly when the working condition changes drastically, and the deviation is restrained quickly; and the fixed PID is responsible for fine adjustment after the system is stable, so that ammonia escape is effectively reduced, and the denitration efficiency is ensured.

Description

A fuzzy PID parallel superposition hybrid control system and method for SCR ammonia injection in a multi-fuel combustion boiler Technical Field

[0001] This invention relates to boiler flue gas pollutant control, and more particularly to a parallel superposition and hybrid control system and method for SCR ammonia injection fuzzy PID in multi-fuel combustion boilers. Background Technology

[0002] With increasingly stringent environmental protection regulations, nitrogen oxides (NOx), as one of the major air pollutants, have become a key challenge for coal-fired power plants, industrial boilers, and other industries. Selective catalytic reduction (SCR) technology, due to its high denitrification efficiency and mature technology, is widely considered the most effective and mainstream technology for controlling NOx emissions from boiler flue gas. The core principle of SCR technology lies in the precise control of the injection amount of the reducing agent (usually ammonia NH3 or urea) under the action of a catalyst, causing NH3 to selectively react with NOx in the flue gas, producing harmless nitrogen (N2) and water (H2O).

[0003] In the SCR denitrification process, the control of ammonia injection rate is crucial. Insufficient ammonia injection (under-injection) will lead to low NOx reduction efficiency, resulting in NOx concentrations that fail to meet environmental standards. Conversely, excessive ammonia injection (over-injection) will cause unreacted ammonia to be released with the flue gas, resulting in "ammonia escape." Ammonia escape not only wastes reducing agent and increases operating costs, but more seriously, the escaped ammonia reacts with SO3 in the flue gas to form ammonium bisulfate (ABS), which easily deposits, clogs, and corrodes the surfaces of downstream equipment such as air preheaters, affecting the safe and economical operation of the boiler. Furthermore, ammonia escape itself is a secondary pollutant. Therefore, achieving precise control of ammonia injection rate to ensure efficient NOx removal while minimizing ammonia escape is key to optimizing the operation of the SCR system.

[0004] However, in practical industrial applications, especially for boilers employing multi-fuel combustion strategies, the operating conditions become exceptionally complex and dynamically variable. These boilers may burn or blend various fuels with vastly different properties, depending on fuel availability, cost-effectiveness, or environmental requirements. Examples include low-calorific-value, highly variable blast furnace gas (BFG), high-calorific-value but similarly complex coke oven gas (COG), liquid fuels such as heavy oil, diesel, or emerging emulsions (such as coal-water slurry, water-in-oil emulsion), and solid fuels such as different types of coal, biomass (such as straw, sawdust, rice husks), and municipal solid waste derived fuels (RDF). These different types of fuels exhibit significant differences in their physicochemical properties (such as calorific value, nitrogen content, moisture, ash content, volatile matter, and combustion rate). When boilers switch between these fuels or mix them in different proportions, the following series of significant dynamic changes occur:

[0005] (1) Dramatic fluctuations in the initial NOx concentration: The nitrogen content and combustion characteristics of different fuels directly affect the flame temperature and excess air coefficient, which in turn leads to a large-scale, even several-order-of-magnitude, change in the initial NOx concentration at the SCR system inlet in a short period of time. For example, the NOx concentration may be high when burning coke oven gas with high nitrogen content or certain biomass, while it may be relatively low when burning blast furnace gas, but the flue gas volume is huge.

[0006] (2) Significant changes in flue gas volume and composition: The total volumetric flow rate of flue gas produced by the combustion of different fuels, as well as the concentrations of components such as CO2, H2O, and O2, will change accordingly, affecting the stoichiometric relationship and heat and mass transfer characteristics of the SCR reaction.

[0007] (3) Flue gas temperature fluctuation: Changes in fuel calorific value and combustion organization will cause fluctuations in furnace temperature and SCR inlet flue gas temperature, which directly affect the catalyst's activity window and reaction rate.

[0008] (4) Complex response characteristics: The response process of fuel switching or co-firing ratio adjustment often has nonlinear, large hysteresis and time-varying characteristics, and the dynamic characteristics of the system under different fuel combinations may be completely different.

[0009] Traditional PID controllers, whose control parameters (proportional Kp, integral Ki, derivative Kd) are typically fixed after tuning or can only be adjusted in simple segments, are ill-suited to the extremely complex operating conditions resulting from the use of various fuels, such as blast furnace gas, coke oven gas, fuel oil, emulsions, and biomass, and their arbitrary combinations. When faced with rapid fuel switching or drastic changes in the co-firing ratio, fixed-parameter PID controllers often exhibit slow response and insufficient adjustment accuracy. Particularly with significant changes in operating conditions, they are prone to under- or over-adjustment of the control output, leading to problems such as excessive ammonia slip or severely insufficient NOx reduction efficiency. This makes it difficult for the system to consistently and stably meet both stringent environmental emission limits and economic operating indicators, especially given the increasingly stringent environmental regulations on NOx emission limits and ammonia slip control.

[0010] To overcome the limitations of traditional PID control in handling the dynamic characteristics of complex systems, existing research has attempted to introduce advanced control algorithms, such as fuzzy control, neural network control, and model predictive control. Among these, fuzzy control, due to its advantages of not relying on an accurate mathematical model of the controlled object, strong robustness, and ability to effectively integrate expert experience, has shown potential in handling nonlinear and time-varying system control problems. The fuzzy PID hybrid control strategy, which combines fuzzy control with PID control, aims to improve the overall performance of the control system by utilizing the intelligent reasoning capabilities of fuzzy logic.

[0011] Therefore, for multi-fuel combustion boilers that burn various fuels or mixtures thereof, such as blast furnace gas, coke oven gas, fuel oil, emulsion, and biomass, there is an urgent practical need and significant engineering application value in developing an advanced control system that can quickly respond to large deviations in the system caused by drastic changes in fuel characteristics and load fluctuations, and work in conjunction with a PID controller to improve the accuracy of SCR ammonia injection control and the robustness of the system. Summary of the Invention

[0012] Purpose of the Invention: The purpose of this invention is to provide a parallel superimposed hybrid control system and method for SCR ammonia injection in multi-fuel combustion boilers. This system aims to overcome the problems of poor adaptability and slow control response caused by the fixed parameters of traditional fixed-parameter PID controllers in SCR ammonia injection control applications of multi-fuel combustion boilers (especially those burning or co-burning various differentiated fuels such as blast furnace gas, coke oven gas, fuel oil, emulsion, and biomass). These problems result in ammonia escape or low NOx reduction efficiency.

[0013] Technical solution: A parallel superposition and hybrid control system for SCR ammonia injection in a multi-fuel combustion boiler, comprising:

[0014] Parameter Monitoring Module: This module is responsible for real-time, continuous, and high-precision monitoring and acquisition of key parameters of the SCR denitrification system. The monitored parameters include at least the flue gas outlet NOx concentration and / or ammonia slip concentration, reflecting the denitrification effect and ammonia consumption. Preferably, this module can also monitor other important parameters affecting the SCR reaction process and control decisions, such as the boiler's current operating load, SCR reactor inlet flue gas temperature, SCR reactor inlet NOx concentration, and flue gas flow rate. These parameters provide real-time operating condition information for subsequent control decisions.

[0015] Deviation and Deviation Change Rate Calculation Unit: This unit receives the target setpoint (e.g., the desired NOx concentration limit at the flue gas outlet). ) and the real-time measured values ​​of the corresponding controlled parameters fed back by the parameter monitoring module (e.g. ), calculate the deviation of the current control cycle and the rate of change of deviation ,in To control the cycle.

[0016] Fixed-parameter PID controller module: This module is a standard digital PID controller with a proportional gain K. p Integral gain K i Differential gain K d It is a fixed constant. It receives the deviation from the deviation and deviation change rate calculation unit. Based on the standard PID control algorithm, a feedback control signal is calculated and generated. :

[0017]

[0018] Among them, K p K i K d For pre-tuned fixed parameters, This is the control cycle. The feedback control signal... This represents a baseline ammonia injection flow rate command.

[0019] Fuzzy Logic Controller (FLC) Module: This module works in parallel with the PID controller, providing a compensating control input. Its main function is to adjust the control input based on the current operating state of the SCR system (via deviation). and rate of change of deviation (Reflection), calculate an additional fuzzy control signal. This signal represents the adjustment amount to the baseline ammonia injection flow rate. The fuzzy logic controller module uses the deviation... and rate of change of deviation As input, it contains fuzzification units, a knowledge base (including a fuzzy rule base and a membership function database), fuzzy inference units, and defuzzification units.

[0020] Fuzzification unit: converts precise, numerical input variables and Through a predefined membership function, it is transformed into a linguistic fuzzy variable (e.g., the deviation). Described as the degree of membership of fuzzy sets such as "negative large", "zero", "positive small").

[0021] knowledge base:

[0022] Membership function database: stores input variables ( ) and output variables ( The definition of fuzzy subsets of and their corresponding membership functions.

[0023] The fuzzy rule base contains a series of "IF-THEN" fuzzy control rules based on expert experience, operational practice, or system identification data. These rules describe the magnitude and direction of the fuzzy control signal to be output under different combinations of deviation (e) and deviation change rate (ec). Specifically, the rule is designed to provide a larger compensation signal that can quickly correct deviations when |e| or |ec| is large. When both |e| and |ec| are small, the provided compensation signal Smaller or close to zero. For example, a rule might take the form: "IF deviation e is negative large AND deviation change rate ec is negative medium THEN fuzzy control signal". "To be upright and righteous."

[0024] Fuzzy inference unit: Based on the input fuzzy value and the rules in the fuzzy rule base, it uses certain fuzzy implication relations and composition rules to perform logical inference and obtain one or more fuzzy control outputs.

[0025] Defuzzification unit: Converts the fuzzy output obtained from the fuzzy inference unit into precise, numerical fuzzy control signals. .

[0026] Control signal superposition module: This module receives the feedback control signal output from the fixed-parameter PID controller module. (Reference ammonia injection flow command) and fuzzy control signal output by the fuzzy logic controller module (Ammonia injection flow rate adjustment) The two are superimposed to generate the final total ammonia injection flow rate control signal. :

[0027]

[0028] This module typically also includes amplitude limiting of the total control signal to ensure it remains within a reasonable physical range.

[0029] Ammonia injection execution module: This module is responsible for accurately executing the total control signal output by the control signal superposition module. It typically includes an ammonia supply and metering unit (such as a flow meter), ammonia flow precision regulating valve, an ammonia / air mixer, and an ammonia injection grid (AIG) to uniformly inject ammonia into the flue gas duct. Based on the received overall control signal, the ammonia injection execution module quickly and accurately adjusts the ammonia injection flow rate to ensure that the appropriate amount of ammonia is fully mixed with the flue gas and enters the SCR reactor.

[0030] Furthermore, the data collected by the parameter monitoring module can undergo necessary signal conditioning, such as filtering, calibration, and unit conversion, to improve the reliability and applicability of the data.

[0031] This invention also provides a corresponding parallel superposition hybrid control method for SCR ammonia injection in multi-fuel combustion boilers. This method is implemented through the above-mentioned system and includes the following steps:

[0032] S1. Real-time parameter monitoring and data acquisition: During the operation of the control system, the parameter monitoring module continuously monitors the operating parameters of the SCR denitrification system in real time, at least obtaining the actual value of the NOx concentration at the flue gas outlet. ) and / or the actual value of ammonia slip concentration ( Simultaneously, boiler load (L) and SCR inlet flue gas temperature (T) can be selectively collected. in ), SCR inlet NOx concentration ( ), flue gas flow rate (F gas Auxiliary control parameters such as ( ).

[0033] S2. Deviation and Deviation Rate of Change Calculation: The control system (usually within the deviation and deviation rate of change calculation unit) calculates the deviation and deviation rate of change based on the preset target setpoint of the controlled parameter (e.g., ...). ) and the actual measured values ​​of the corresponding parameters obtained in step S1 (e.g. ), calculate the deviation of the current control cycle At the same time, calculate the current deviation. Deviation relative to the previous control cycle rate of change ,in, To control the cycle.

[0034] S3. Fixed-parameter PID control calculation: The fixed-parameter PID controller module receives the deviation calculated in step S2. Using a pre-tuned fixed parameter K p K i K d and control cycle The feedback control signal for this control cycle is calculated based on the PID control algorithm (as described in the aforementioned formula). This serves as the baseline ammonia injection flow rate instruction.

[0035] S4. Fuzzy logic compensation calculation: The deviation calculated in step S2 and rate of change of deviation It is transmitted as an input signal to the fuzzy logic controller module.

[0036] S41. Fuzzification: The fuzzification unit of the fuzzy logic controller module uses a predefined membership function to... and The precise values ​​are respectively converted into fuzzy linguistic variables and their membership degrees in their respective fuzzy domains.

[0037] S42, Fuzzy Reasoning: The fuzzy reasoning unit (103c) bases its reasoning on the fuzzified input and the fuzzy rule base stored in the knowledge base (103b) (e.g., a series of "IF e is A AND ec is B THEN"). (using the "is C" rule) to perform fuzzy logic operations, resulting in a fuzzy output representing the fuzzy control signal.

[0038] S43. Defuzzification: The defuzzification unit converts the fuzzy output obtained from fuzzy inference into a precise, numerical fuzzy control signal. This serves as the adjustment amount for the ammonia injection flow rate. The rule and defuzzification design aims to ensure that when |e| or |ec| is large, | When |e| and |ec| are relatively large, rapid coarse adjustment is achieved; when both |e| and |ec| are relatively small, Smaller values ​​or values ​​close to zero allow for fine-tuning of the PID controller.

[0039] S5. Control Signal Overlay and Limiting: The control signal overlay module overlays the baseline ammonia injection flow command calculated in step S3. The ammonia injection flow rate adjustment calculated in step S4 The signals are superimposed to obtain the total ammonia injection flow control signal. Then to Amplitude limiting is performed to obtain the final control signal. .

[0040] S6. Ammonia Injection Quantity Adjustment Execution: The ammonia injection execution module receives the final control signal from the control signal superposition module. Based on this, the opening of the ammonia flow regulating valve or other control elements are precisely adjusted, thereby changing the actual amount of ammonia injected into the flue gas duct of the SCR reactor.

[0041] S7. Cyclic Control: The system returns to step S2 and enters the next control cycle, continuing the above closed-loop control process to dynamically adapt to changes in operating conditions and maintain the efficient and stable operation of the SCR system.

[0042] Beneficial effects:

[0043] (1) Significantly improves response speed and control performance to drastic changes in operating conditions: When drastic changes in fuel characteristics (such as fuel switching) or large fluctuations in boiler load cause drastic changes in flue gas parameters (NOx concentration, flue gas velocity, temperature, etc.), resulting in large deviations |e| and deviation change rates |ec| in the system, the fuzzy logic controller can quickly generate a significant compensation control signal. The compensation signal, superimposed on the output of the fixed-parameter PID controller, can provide a rapid and significant adjustment of the ammonia quantity, quickly bringing the outlet NOx concentration closer to the target value. This effectively addresses such strong disturbance conditions, shortens the time it takes for the system to reach stability, and achieves rapid "coarse adjustment."

[0044] (2) Improved control accuracy and effective reduction of ammonia slip: The rapid coarse adjustment of fuzzy logic combined with the fine adjustment of fixed parameter PID when approaching the target value enables the system to achieve more accurate ammonia control whether in dynamic changes or in a steady state. Especially when dealing with fuel switching, the rapid compensation of fuzzy control avoids the lag and under-adjustment that may occur with traditional fixed parameter PID, thereby reducing NOx exceedances caused by insufficient response; at the same time, after the system approaches stability, the fuzzy compensation signal decreases, and fine adjustment is mainly performed by fixed PID, which can effectively avoid excessive ammonia injection caused by over-compensation, thereby strictly controlling the ammonia slip concentration at an extremely low level.

[0045] (3) Enhanced robustness and adaptability to multiple fuels in the control system: Fuzzy control does not rely on a precise mathematical model of the controlled object, but incorporates experience in dealing with complex nonlinear behavior and uncertainties (especially disturbances introduced by unknown or rapid changes in fuel composition) through a rule base. Unlike fixed-parameter PID controllers that directly act on the controlled object, fuzzy controllers provide a compensation signal based on deviation and the rate of change of deviation. This parallel superposition structure has a strong tolerance for the uncertainty of model parameters. The system has a stronger resistance to unknown disturbances that may occur under various fuel combinations and load conditions, and can still maintain good control performance and stability even when fuel characteristics undergo large unknown changes.

[0046] (4) Reduce the operating cost of the SCR system and optimize the economy of multi-fuel combustion: Precise control of ammonia consumption, especially by avoiding under-adjustment through rapid response and reducing over-adjustment through precise control, comprehensively reduces the consumption of reducing agent and directly saves procurement costs. Effective control of ammonia escape reduces deposition, blockage and corrosion of downstream equipment, and reduces maintenance costs. Stable NOx emissions that meet standards avoid environmental penalties. These measures together improve the economic efficiency of boiler operation.

[0047] (5) Easy to implement in engineering and adjust parameters, adaptable to different fuel characteristics: The tuning of fixed-parameter PID is relatively mature and can be performed for nominal operating conditions. Although the design of fuzzy control rules requires experience, its inputs are deviation and deviation change rate, and the output is compensation quantity, which is relatively intuitive. It can be adjusted and optimized by combining the experience of operators for the combustion characteristics and SCR response characteristics of different fuels (such as blast furnace gas, coke oven gas, fuel oil, biomass, etc.). The system structure is clear, and fuzzy control modules and superposition units can be added to the existing PID control system to achieve upgrades and transformations, which has good engineering practicality and promotion value. Attached Figure Description

[0048] Figure 1 is a block diagram of the SCR ammonia injection fuzzy PID parallel superposition hybrid control system for a multi-fuel combustion boiler.

[0049] Figure 2 is a schematic diagram of the internal structure of the fuzzy logic controller module (FLC);

[0050] Figure 3 is a flowchart of the parallel superposition and hybrid control method of SCR ammonia injection fuzzy PID in multi-fuel combustion boiler. Detailed Implementation

[0051] To make the technical solution of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0052] Example 1

[0053] As shown in Figure 1, this embodiment discloses a fuzzy PID parallel superposition hybrid control system for boiler flue gas SCR denitrification devices used in multi-fuel combustion (e.g., combustion of multiple coal types or co-firing of coal and biomass, especially including co-firing of blast furnace gas, coke oven gas, fuel oil, emulsion, etc.). This system aims to adjust the ammonia injection rate through the coordinated operation of a fixed-parameter PID controller and a parallel fuzzy controller, effectively addressing the impact of fuel variations and boiler load fluctuations on the denitrification process, and achieving continuous compliance with NOx emission standards and strict control of ammonia slip. The system mainly consists of the following cooperative modules:

[0054] Parameter monitoring module 101:

[0055] This module is responsible for comprehensively and in real-time monitoring the operational status of the SCR system. Specifically:

[0056] A NOx analyzer is installed at the flue gas outlet of the SCR reactor to continuously monitor the NOx concentration in the outlet flue gas. .

[0057] An ammonia slip analyzer is installed at the SCR reactor outlet or at a suitable location downstream to monitor the ammonia slip concentration in real time. (For example, the target is to keep it below 3 ppm).

[0058] To obtain more comprehensive operating condition information to support control decisions, especially when dealing with drastic changes in operating conditions caused by burning multiple fuels such as blast furnace gas, coke oven gas, fuel oil, emulsions, and biomass, this module can also integrate:

[0059] Boiler load sensor / signal interface: Acquires boiler load level L (e.g., range 30%-100%).

[0060] Flue gas temperature sensor: monitors the inlet flue gas temperature T of the SCR reactor. in For example, for medium- and low-temperature catalysts, this temperature is typically controlled between 280°C and 400°C.

[0061] Inlet NOx concentration sensor: monitors the initial NOx concentration before entering the SCR reactor. For example, when using blast furnace gas, Possibly 100-250 mg / Nm 3 When switching to co-firing a high proportion of coke oven gas or certain high-nitrogen biomass, It may rise sharply to 800-1500 mg / Nm 3 Or higher.

[0062] Flue gas flow meter: monitors the flow rate F of flue gas passing through the SCR reactor. gas For example, the full-load flue gas volume of a 350MW boiler may be between 1,200,000 and 1,500,000 Nm³. 3 / h (dry basis) range.

[0063] All monitoring signals, after signal conditioning (such as filtering, with a sampling period of 1 second), are transmitted to the subsequent control module. The target controlled parameter is typically... Its target set value It can be set by the upper-level system or preset according to environmental protection requirements; for example, it can be set to 40 mg / Nm³. 3 (Dry basis, corresponding to standard oxygen content, such as 6%).

[0064] Deviation and Deviation Change Rate Calculation Unit 105:

[0065] As shown in Figure 1, this unit (node ​​105 in Figure 1) receives... (e.g., 40 mg / Nm³) and the parameters collected by the parameter monitoring module 101 Calculate the current deviation Simultaneously calculate the rate of change of deviation. ,in To control the cycle (e.g., (seconds). Calculation results and The data is transmitted to the fixed-parameter PID controller module 102 and the fuzzy logic controller module 103, respectively.

[0066] Fixed parameter PID controller module 102:

[0067] The core of this module is a digital PID controller with a proportional gain K. p Integral gain K i Differential gain K d It is a constant that is pre-tuned and fixed.

[0068] Input: Deviation .

[0069] Processing: Based on a fixed K p K i K d The parameters are used to calculate the feedback control signal using a standard PID algorithm. .

[0070]

[0071] (This requires engineering treatments such as integral saturation prevention, for example, limiting the output of the integral term to within ±30% of the total output).

[0072] For example, for the aforementioned 350MW boiler under a certain reference fuel and load, the PID parameters initially tuned through simulation or field tests can be: K p =0.6, K i ==0.3, K d =0.15. These parameters are in the PID output. Represents a standardized control signal (e.g., 0-1 or 0-100%), and the deviation... These values ​​are determined by standardization or by including the unit effect in the gain coefficient. These parameter values ​​can be adjusted based on actual operating experience and the specific physical meaning of the control signals.

[0073] Output: Feedback control signal This signal represents a baseline ammonia injection flow command (e.g., in kg / h, the specific value of which is determined by PID parameters and deviation) and is transmitted to the control signal superposition module 106.

[0074] Fuzzy Logic Controller (FLC) Module 103:

[0075] This is a compensation component that works in parallel with a fixed-parameter PID controller. As shown in Figure 2, the detailed structure and working principle of the FLC module 103 are as follows:

[0076] Input variables: Receives the deviation from the deviation and deviation change rate calculation unit 105. and rate of change of deviation .

[0077] Fuzzification unit 103a: for input variables and Blur the image.

[0078] The universe of discourse for deviation e: for example, according to For 40 mg / Nm³ and its possible fluctuations, it is set as [-30 mg / Nm³, +30 mg / Nm³].

[0079] The domain of the rate of change of deviation ec: for example, set as [-10 mg / Nm³ / s, +10 mg / Nm³ / s].

[0080] The universe of discourse for each input variable is divided into 7 fuzzy subsets: NB (negative large), NM (negative medium), NS (negative small), ZO (zero), PS (positive small), PM (positive medium), and PB (positive large).

[0081] Membership functions: For example, triangular membership functions are used. Taking deviation e as an example, the vertex of the membership function of its fuzzy subset "ZO" is defined at e=0mg / Nm³, and the base of the triangle covers the range of [-5 mg / Nm³, +5 mg / Nm³]; the vertex of the fuzzy subset "NS" is defined at e=-10mg / Nm³, and the base covers the range of [-15 mg / Nm³, -5 mg / Nm³]; the membership functions of other fuzzy subsets are similarly distributed symmetrically or asymmetrically within the universe of discourse.

[0082] Knowledge Base 103b:

[0083] Database (Membership Function Library): Stores input variables e, ec and output variables (fuzzy control signals). The above-defined fuzzy subsets and membership function parameters. Output variables. This represents the adjustment amount to the baseline ammonia injection flow rate, and its domain can be set, for example, to a specific ammonia injection flow rate adjustment range, such as [-5 kg / h, +5 kg / h]. Its fuzzy subset can also be divided into seven levels, from NB to PB.

[0084] The rule base contains a set of 49 rules as shown in Table 1: "IF e is A AND ec is B THEN". Fuzzy control rules of the form "is C" are designed to provide compensation signals. The key lies in the design rules, ensuring that when |e| or |ec| is large (indicating drastic changes in operating conditions causing the system to deviate significantly from the target), It can provide a large compensation amount that can quickly correct deviations; when both |e| and |ec| are small (the system is close to stable or under normal operating conditions), Smaller or close to zero.

[0085] Table 1 Fuzzy Rule Table

[0086] e\ecNBNMNSZOPSPMPBPBZONSNMNBNBNBNBPMPSZONSNMNMNBNBPSPMPSZONSNNMNMNBZOPBPMPSZONSNMNBNSPBPMPMPSZOPSPMNMPBPBPMPMPSZOPSNBPBPBPBPPMPSZO surface

[0087] For example:

[0088] Rule 1: IF e is NB (Exported NOx far exceeds the standard, e = -30mg / Nm³, corresponding to...) mg / Nm³) AND ec is NB (and is still deteriorating rapidly, e.g., ec = -8 mg / Nm³ / s) THEN is PB (outputs a large positive compensation signal, for example) (This corresponds to an additional +4 kg / h ammonia flow rate). This rule is particularly applicable to the initial stage when fuel is suddenly switched from low-NOx fuel to high-NOx fuel.

[0089] Rule 2: IF e is PB (Export NOx is significantly lower than the target, possibly indicating excessive ammonia injection; e = +30 mg / Nm³, corresponding to...) mg / Nm³) AND ec is ZO (not much change) THEN is NB (outputs a large negative compensation signal, for example) This corresponds to an additional reduction of -4 kg / h of ammonia flow. This may occur when switching from a high-load, high-NOx operating condition to a low-load, low-NOx operating condition, or from a high-NOx fuel to a low-NOx fuel, where system response lag leads to excess ammonia injection, requiring a rapid reduction in ammonia flow.

[0090] Rule 3: IF e is ZO AND ec is ZO THEN is ZO (output compensation signal is zero, i.e.) (Adjustment of ammonia flow rate corresponding to 0 kg / h).

[0091] Fuzzy reasoning unit 103c: performs fuzzy reasoning based on input membership degree and rule base, for example, using Mamdani-type reasoning method (minimum implication, maximum composition).

[0092] Defuzzification unit 103d: Converts the fuzzy output obtained from fuzzy inference into precise numerical fuzzy control signals. For example, the commonly used Center of Gravity (CoG) method can be employed.

[0093] Control signal superposition module 106:

[0094] This module receives the reference ammonia injection flow rate command output by the fixed-parameter PID controller module 102. (Unit: kg / h) and the ammonia injection flow rate adjustment output by the fuzzy logic controller module 103 (Unit: kg / h)

[0095] Processing: Calculate the total ammonia injection flow control signal :

[0096]

[0097] Then to Limiting is performed (e.g., ensuring the total ammonia injection rate is between 0 kg / h and the equipment's maximum ammonia injection capacity, such as 80 kg / h, and the rate of change does not exceed the allowable value) to obtain the final control signal. .

[0098] Output: Final control signal (Unit: kg / h) is transmitted to the ammonia injection execution module 104.

[0099] Ammonia injection execution module 104:

[0100] This module receives the final control signal. This allows for precise adjustment of the ammonia flow rate. For example, by controlling the opening degree of the ammonia flow regulating valve.

[0101] The specific flow of the control method is shown in Figure 3:

[0102] Step S1: System Initialization and Parameter Monitoring Startup. The control system starts and initializes the fixed PID parameters (K... p K i K d (e.g., in the example above), the membership function and rule base of the fuzzy controller (e.g., in the example above and Table 1), etc. Set the control cycle. Seconds. Parameter monitoring module 101 begins continuously collecting NOx concentration data at the SCR outlet. Other relevant parameters. Set the target outlet NOx concentration. (e.g., 40 mg / Nm³).

[0103] Step S2: Calculate the deviation e(k) and the rate of change of deviation ec(k). In each control cycle k, the following calculations are performed using calculation unit 105:

[0104]

[0105]

[0106] Step S3: Fixed-parameter PID control calculation. The fixed-parameter PID controller module 102 receives e(k) and uses a fixed K. p K i K d and control cycle Calculate the baseline ammonia injection flow command .

[0107] Step S4: Fuzzy logic compensation calculation. The fuzzy logic controller module 103 receives e(k) and ec(k). Through fuzzification, fuzzy inference, and defuzzification processes (as described in the FLC module above with specific parameters and methods), the ammonia injection flow adjustment amount for this cycle is calculated. .

[0108] Step S5: Control signal superposition and limiting. The control signal superposition module 106 receives... and Calculate the total ammonia injection flow control signal .right Amplitude limiting is performed to obtain .

[0109] Step S6: Perform ammonia injection adjustment. Ammonia injection execution module 104 receives... And accordingly, the ammonia flow rate is precisely adjusted.

[0110] Step S7: Loop and Feedback. The system waits for the next sampling moment (1 second later), then returns to step S2, forming a continuous closed-loop feedback control.

[0111] Suppose a multi-fuel boiler (such as the 350MW boiler mentioned above) is stably burning blast furnace gas (BFG). Approximately 150 mg / Nm³ The ammonia injection flow rate stabilizes around 40 mg / Nm³, at which point the PID controller outputs the baseline ammonia injection flow rate. It may be 25 kg / h, the output of FLC. The flow rate was 0 kg / h. Suddenly, the system switched to co-firing a high proportion of coke oven gas (COG) and a small amount of biomass, resulting in... It rapidly climbed to 1200 mg / Nm³ within minutes.

[0112] Dramatic parameter change: Parameter monitoring module 101 detected A rapid increase, for example, from 40 mg / Nm³ to 70 mg / Nm³, can lead to deviations. It quickly becomes a very large negative value (e.g., mg / Nm³, corresponding to the fuzzy set NB). Since the outlet NOx concentration is still rising rapidly, the deviation change rate... It also shows a significant negative rapid change (e.g., after continuous sampling and computation, It may reach -8 mg / Nm³ / s, corresponding to the fuzzy set NB, and this value is within the set universe of discourse [-10 mg / Nm³ / s, +10 mg / Nm³ / s].

[0113] FLC Response (Coarse Adjustment): FLC module 103 received these dramatic changes. and Signal. Based on its fuzzy rules optimized for such abrupt changes in operating conditions caused by drastic changes in fuel characteristics (such as from BFG to COG+biomass) (e.g., "IF e is NB AND ec is NB THEN") When "is PB"), the FLC will quickly output a large positive compensation signal. (For example, this corresponds to an additional ammonia flow rate of +4 kg / h).

[0114] PID Action (Fine-tuning): Simultaneously, the fixed-parameter PID controller module 102 also receives a large negative deviation. It is based on its fixed K. p K i K d The parameters calculate the reference ammonia injection flow command output by the parameter calculation. This value will also gradually increase as the deviation persists (e.g., starting from 25 kg / h).

[0115] Signal superposition and execution: The control signal superposition module 106 superimposes the larger positive compensation signal output by the FLC. (e.g., +4 kg / h) and fixed PID output (For example, it may have been adjusted to 30 kg / h at this point) Adding these together results in a significantly increased total control signal. (For example, 30 + 4 = 34 kg / h). Based on this increased signal, the ammonia injection execution module 104 rapidly and significantly increases the ammonia injection rate.

[0116] Cooperative control effect: Due to the rapid compensation effect (coarse adjustment) of the fuzzy controller, the rate and magnitude of the increase in ammonia concentration are significantly higher than those achieved with only a fixed PID controller. This allows for faster suppression of the upward trend in outlet NOx concentration, quickly pulling it back to near the target value. When the outlet NOx concentration approaches the target value (e.g., e becomes NS or ZO, ec becomes ZO), the deviation... and rate of change of deviation The output of FLC decreases. This also decreases and approaches zero (e.g., becomes 0 kg / h). At this point, the total control signal is mainly generated by the output of a fixed-parameter PID controller. The system is composed of a fixed PID controller that performs precise fine-tuning around the target value to eliminate steady-state errors and maintain system stability. This collaborative approach of "coarse tuning first, fine tuning later" enables the system to exhibit better dynamic response and control accuracy when facing drastic changes in operating conditions caused by multi-fuel combustion.

[0117] The fuzzy rule base, membership function, and PID parameters of this invention can be continuously improved and adjusted through offline simulation optimization, combined with the operating data of specific boilers and expert experience, to adapt to the specific characteristics of specific boilers and SCR systems, thereby achieving better control effects.

[0118] Although this embodiment primarily uses the outlet NOx concentration as the controlled variable, those skilled in the art will understand that ammonia slip concentration can also be used as the primary controlled variable, or a multivariate control strategy can be employed to consider both. Furthermore, in addition to e and ec, boiler load, flue gas temperature, and other auxiliary inputs can be introduced into the fuzzy controller to form a more complex but potentially more precise control structure.

[0119] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A parallel superposition and hybrid control system for SCR ammonia injection in a multi-fuel combustion boiler, characterized in that, include: The parameter monitoring module (101) is used to monitor the operating parameters of the SCR denitrification system in real time, including at least the NOx concentration and / or ammonia slip concentration at the flue gas outlet; the deviation and deviation change rate calculation unit (105) is used to calculate the deviation e(k) and deviation change rate ec(k) based on the target setpoint and the real-time measurement value monitored by the parameter monitoring module (101); the fixed parameter PID controller module (102) is used to generate a reference ammonia injection flow command U based on the deviation e(k). PID (k); Fuzzy logic controller module (103), used to generate ammonia injection flow rate adjustment U based on the deviation e(k) and the deviation change rate ec(k). FLC (k); Control signal superposition module (106), used to superimpose the reference ammonia injection flow command U PID (k) and the ammonia injection flow rate adjustment amount U FLC (k) superposition generates the total ammonia injection flow control signal U. total (k); Ammonia injection execution module (104), used to control the total ammonia injection flow rate according to the total ammonia injection flow rate control signal U total (k) Adjust the amount of ammonia injected.

2. A parallel superposition hybrid control method for SCR ammonia injection fuzzy PID control in a multi-fuel combustion boiler, characterized in that, Includes the following steps: S1. Monitor the operating parameters of the SCR denitrification system in real time, and obtain at least the actual value of NOx concentration at the flue gas outlet, C. NOx_out S2, Calculate the deviation e(k) = C NOx_setpoint - C NOx_out (k) and the deviation change rate ec(k) = (e(k) - e(k-1)) / ΔT, where ΔT is the control cycle; S3, the reference ammonia injection flow command U is calculated by the fixed parameter PID controller based on the deviation e(k). PID (k); S4. Calculate the ammonia injection flow rate adjustment U based on the deviation e(k) and the deviation change rate ec(k) using a fuzzy logic controller. FLC (k); S5, U PID (k) and U FLC (k) Superimpose to generate total ammonia injection flow control signal U total (k), and for U total (k) Perform amplitude limiting processing to obtain the final control signal U final (k); S6, according to U final (k) Adjust the ammonia injection rate; S7. Repeat steps S2 to S6 to achieve closed-loop control.

3. The SCR ammonia injection fuzzy PID parallel superposition hybrid control system for multi-fuel combustion boilers according to claim 1, characterized in that, The parameter monitoring module (101) is also used to monitor at least one of the following parameters: boiler load, flue gas temperature at the SCR reactor inlet, NOx concentration at the SCR reactor inlet, and flue gas flow rate.

4. The SCR ammonia injection fuzzy PID parallel superposition hybrid control system for multi-fuel combustion boilers according to claim 1, characterized in that, The fuzzy logic controller module (103) includes a fuzzification unit (103a), a knowledge base (103b), a fuzzy reasoning unit (103c), and a defuzzification unit (103d), wherein the knowledge base (103b) stores membership functions and a fuzzy rule base.

5. The SCR ammonia injection fuzzy PID parallel superposition hybrid control system for multi-fuel combustion boilers according to claim 1, characterized in that, The fuzzy rule base contains fuzzy control rules based on deviation e and deviation change rate ec. These fuzzy control rules are configured to output a larger U when |e| or |ec| is larger. FLC (k) to quickly correct deviations; when |e| and |ec| are small, the output U is smaller. FLC (k) or zero.

6. The SCR ammonia injection fuzzy PID parallel superposition hybrid control system for multi-fuel combustion boilers according to claim 1, characterized in that, The control signal superposition module (106) also controls the total ammonia injection flow rate U. total (k) Limit the amount of ammonia injected to ensure that the amount of ammonia injected is within the physically permissible range.

7. The multi-fuel combustion boiler SCR ammonia injection fuzzy PID parallel superposition hybrid control method according to claim 2, characterized in that, In step S4, the ammonia injection flow rate adjustment amount U is calculated using a fuzzy logic controller. FLC (k) includes the following sub-steps: S41, fuzzification, converting the deviation e(k) and the rate of change of deviation ec(k) into fuzzy variables; S42, fuzzy inference, performing logical inference based on the fuzzy rule base; S43, defuzzification, converting the fuzzy output into a precise U... FLC (k).

8. The multi-fuel combustion boiler SCR ammonia injection fuzzy PID parallel superposition hybrid control method according to claim 2, characterized in that, The fuzzy rules of the fuzzy logic controller are based on the deviation e and the deviation change rate ec. The rules are designed to provide significant compensation when the system deviation is large or changes rapidly, and to reduce the compensation when the system is close to stability.

9. The multi-fuel combustion boiler SCR ammonia injection fuzzy PID parallel superposition hybrid control method according to claim 2, characterized in that, The limiting process in step S5 includes limiting U total (k) Limit the ammonia injection rate between the minimum and maximum ammonia injection rates and control the rate of change.

10. The SCR ammonia injection fuzzy PID parallel superposition hybrid control system for multi-fuel combustion boilers according to claim 1, characterized in that, The system is applicable to boilers that burn or co-fire multiple fuels, including at least one of blast furnace gas, coke oven gas, fuel oil, emulsion, and biomass.

Images