Industrial boiler combustion state real-time monitoring and energy efficiency optimization system based on internet of things

Abstract

This invention relates to an IoT-based real-time monitoring and energy efficiency optimization system for industrial boiler combustion status, belonging to the field of industrial automation control technology. It includes: a data acquisition module for acquiring physical sensor data and fuel characteristic data of the boiler, including flue gas oxygen concentration and fuel characteristic data including fuel calorific value and fuel moisture content; a state estimation module for estimating the instantaneous effective excess air coefficient characterizing the current combustion state based on the physical sensor data and fuel characteristic data acquired by the data acquisition module; and a predictive control module for calculating a feedforward control adjustment based on the instantaneous effective excess air coefficient determined by the state estimation module and a preset target excess air coefficient. This invention solves the problem of delayed results in existing technologies that can only passively respond.

Description

Technical Field

[0001] This invention relates to the field of industrial automation control technology, specifically to an Internet of Things-based real-time monitoring and energy efficiency optimization system for industrial boiler combustion status. Background Technology

[0002] As a core energy conversion device, the combustion efficiency and operational stability of industrial boilers have a decisive impact on overall production costs and pollutant emissions. Existing boiler combustion control systems typically employ feedback control logic based on physical sensors. For example, based on the oxygen content sensor readings in the flue gas, a proportional-integral-derivative controller adjusts the air supply. This reactive control paradigm results in control commands being generated with a lag behind actual changes in the combustion state. When faced with real-time fluctuations in fuel components such as calorific value, moisture, and volatile matter, or nonlinear changes in operating load, this lag prevents the control system from anticipating and responding to deviations in the combustion state. Consequently, the system must passively balance incomplete combustion (leading to excessive carbon monoxide and reduced thermal efficiency) with excessive air supply (leading to increased flue gas heat loss and increased nitrogen oxide generation). This fails to eliminate the dynamic conflict between complete combustion and thermal efficiency, hindering further improvements in the overall boiler energy efficiency.

[0003] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0004] The purpose of this invention is to provide an Internet of Things-based real-time monitoring and energy efficiency optimization system for the combustion status of industrial boilers, in order to solve the problems mentioned in the background art.

[0005] The technical solution of the present invention includes: a data acquisition module, used to acquire physical sensor data and fuel characteristic data of the boiler, wherein the physical sensor data includes flue gas oxygen concentration and the fuel characteristic data includes fuel calorific value and fuel moisture content;

[0006] The state estimation module is used to estimate the instantaneous effective excess air coefficient, which characterizes the current combustion state, based on physical sensor data and fuel characteristic data acquired by the data acquisition module.

[0007] The predictive control module is used to calculate the feedforward control adjustment based on the instantaneous effective excess air coefficient determined by the state estimation module and the preset target excess air coefficient.

[0008] The instruction fusion module is used to fuse the feedforward control adjustment calculated by the predictive control module with the corrective control instruction generated by the feedback control loop to determine the final control instruction in order to regulate the combustion process of the boiler.

[0009] Preferably, the state estimation module is used for:

[0010] Determine the model prediction compensation term; determine the apparent excess air coefficient based on the flue gas oxygen concentration; combine the model prediction compensation term, the apparent excess air coefficient, and the estimated value of the instantaneous effective excess air coefficient at the previous moment to iteratively update and generate the instantaneous effective excess air coefficient at the current moment.

[0011] Preferably, determining the model prediction compensation term includes:

[0012] The calorific value deviation is determined based on the difference between the fuel calorific value and the preset benchmark calorific value; the moisture deviation is determined based on the difference between the fuel moisture content and the preset benchmark moisture content; the calorific value deviation is multiplied by the preset calorific value influence coefficient to obtain the first product; the moisture deviation is multiplied by the preset moisture influence coefficient to obtain the second product; the first product and the second product are summed to calculate the model prediction compensation term.

[0013] Preferably, the state estimation module is further used for:

[0014] Determine the dynamic fusion gain; sum the estimated instantaneous effective excess air coefficient of the previous moment with the model prediction compensation term to obtain the predicted value of the current moment; calculate the difference between the apparent excess air coefficient and the predicted value, and multiply it by the dynamic fusion gain to obtain the observation correction term; sum the observation correction term with the predicted value to generate the instantaneous effective excess air coefficient of the current moment.

[0015] Preferably, determining the dynamic fusion gain includes:

[0016] When the short-term variance of the apparent excess air coefficient is lower than a preset threshold, a larger preset value is assigned to the dynamic fusion gain.

[0017] When the short-term variance of the apparent excess air coefficient is not lower than a preset threshold, a preset small value is assigned to the dynamic fusion gain.

[0018] Preferably, the predictive control module is used for:

[0019] The prediction deviation is obtained by subtracting the preset target excess air coefficient from the instantaneous effective excess air coefficient; the time change rate of the instantaneous effective excess air coefficient is calculated; and the feedforward control adjustment amount is calculated based on the prediction deviation, the time change rate, the preset proportional feedforward gain, and the preset differential feedforward gain.

[0020] Preferably, the instruction fusion module is used for:

[0021] Determine the feedforward weighting coefficient; multiply the feedforward control adjustment by the feedforward weighting coefficient to obtain the weighted feedforward quantity;

[0022] The weighted feedback quantity is obtained by multiplying the corrective control command by the difference between the feedforward weighting coefficient and the weighted feedforward quantity. The weighted feedforward quantity, the weighted feedback quantity, and the basic control signal determined according to the boiler load are summed to generate the final control command.

[0023] Preferably, determining the feedforward weighting coefficients includes:

[0024] When the absolute value of the model's predicted compensation term or the absolute value of the time-varying rate of change of the instantaneous effective excess air coefficient exceeds the preset disturbance threshold, the feedforward weight coefficient is increased.

[0025] When the absolute value of the model's predicted compensation term and the absolute value of the time-varying rate of change of the instantaneous effective excess air coefficient do not exceed the preset disturbance threshold, the feedforward weight coefficient is reduced.

[0026] This invention provides an improved IoT-based real-time monitoring and energy efficiency optimization system for industrial boiler combustion status, which has the following improvements and advantages compared to existing technologies:

[0027] 1. It achieves proactive compensation for core disturbance sources; by setting up a data acquisition module, this invention not only acquires physical sensor data of the boiler, but more importantly, it also proactively acquires fuel characteristic data that characterizes the disturbance source. This enables the system to pre-calculate the amount of compensation needed before inferior fuel enters the furnace and actually affects combustion, thus fundamentally solving the problem of delayed results caused by the passive response of existing technologies.

[0028] 2. A high-confidence real-time understanding of the internal state of the combustion process is achieved. The state estimation module in this invention employs a core state estimation iterative expression, which is a typical state observer. It is not a simple mathematical derivation, but rather an engineering application definition of control theory. It fuses three pieces of information: the state estimate λ from the previous moment. t-1 The inertia of the system state is represented; the observation correction term uses the measurements of physical sensors to correct the estimate; while the feedforward compensation term injects the prediction of future disturbances; by dynamically adjusting the fusion gain, the system can intelligently weigh the degree of trust in the sensor measurements. The instantaneous effective excess air coefficient estimate output by this module has higher accuracy and smoothness compared to the single, noisy and delayed sensor signals that existing technologies rely on, providing a solid data foundation for subsequent precise control.

[0029] 3. It realizes a transformation from passive correction to active intervention in control mode; the predictive control module calculates the feedforward control adjustment amount based on high-precision state estimation values ​​through feedforward control laws; its control effect is not only proportional to the current deviation, but also includes the prediction of the deviation change trend; when the boiler load suddenly increases, even if it has not deviated significantly from the target value, its rapid downward trend will be captured by the differential term, thereby generating a strong, proactive control effect that increases the airflow in advance; this initiative changes the situation where existing technologies must wait for error accumulation before responding, greatly improving the dynamic response speed and stability of the system;

[0030] 4. It achieves the optimal fusion of control speed and accuracy; the instruction fusion module realizes the dynamic coordination of feedforward and feedback control strategies through the final control instruction; it overcomes the limitations of the single control mode in the prior art, so that the system of the present invention can exhibit the best control performance under any working condition; Attached Figure Description

[0031] The present invention will be further explained below with reference to the accompanying drawings and embodiments:

[0032] Figure 1 This is a flowchart of the system of the present invention. Detailed Implementation

[0033] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0034] Example 1:

[0035] Please see Figure 1 This invention provides an IoT-based real-time monitoring and energy efficiency optimization system for the combustion status of industrial boilers, including: a data acquisition module for acquiring physical sensor data and fuel characteristic data of the boiler, wherein the physical sensor data includes flue gas oxygen concentration and the fuel characteristic data includes fuel calorific value and fuel moisture content;

[0036] The state estimation module is used to estimate the instantaneous effective excess air coefficient, which characterizes the current combustion state, based on physical sensor data and fuel characteristic data acquired by the data acquisition module.

[0037] The predictive control module is used to calculate the feedforward control adjustment based on the instantaneous effective excess air coefficient determined by the state estimation module and the preset target excess air coefficient.

[0038] The instruction fusion module is used to fuse the feedforward control adjustment calculated by the predictive control module with the corrective control instruction generated by the feedback control loop to determine the final control instruction in order to regulate the combustion process of the boiler.

[0039] In this embodiment, the IoT-based industrial boiler combustion status real-time monitoring and energy efficiency optimization system achieves optimized control of the boiler combustion process through the collaborative work of its functional modules. The system's sensing unit, i.e., the data acquisition module, continuously acquires high-frequency physical quantities such as flue gas oxygen concentration (O2) from the boiler flue gas analyzer, and receives batch calorific value (q) and moisture content (w) of the fuel via the IoT interface. t The state estimation module processes this multi-source data stream, and its function is to calculate the internal variables characterizing the core combustion state, namely the instantaneous effective excess air coefficient λ; based on this accurate λ estimate and the optimal target value λ... r The deviation is used to generate a feedforward control adjustment Δu by the predictive control module. f The instruction fusion module will adjust the feedforward amount Δu f The correction command u of the traditional proportional-integral-derivative feedback loop p Perform dynamic weighted fusion and output the final control command u. t This architecture transforms the system control logic from a lagging error response to an advanced disturbance prediction. When the fuel composition or operating load changes, it can maintain the combustion process in the high-efficiency operating range, thus optimizing combustion completeness and thermal efficiency.

[0040] Example 2

[0041] The state estimation module is used for:

[0042] Determine the model prediction compensation term; determine the apparent excess air coefficient based on the flue gas oxygen concentration; combine the model prediction compensation term, the apparent excess air coefficient, and the estimated value of the instantaneous effective excess air coefficient at the previous moment to iteratively update and generate the instantaneous effective excess air coefficient at the current moment.

[0043] Determine the model prediction compensation term, including:

[0044] The calorific value deviation is determined based on the difference between the fuel calorific value and the preset benchmark calorific value; the moisture deviation is determined based on the difference between the fuel moisture content and the preset benchmark moisture content; the calorific value deviation is multiplied by the preset calorific value influence coefficient to obtain the first product; the moisture deviation is multiplied by the preset moisture influence coefficient to obtain the second product; the first product and the second product are summed to calculate the model prediction compensation term.

[0045] In this embodiment, the estimation logic of the state estimation module is based on a two-step calculation process; to cope with real-time changes in the intrinsic properties of the fuel, the module calculates a model prediction compensation term Δλ. cThe underlying logic of this calculation is that fluctuations in fuel calorific value and moisture content are the root causes of deviations in combustion state. By introducing a compensation term based on fuel characteristic deviations, the system can proactively compensate for these fluctuations at the input before they substantially affect the combustion process. The composition of this compensation term is given by a linear approximation model.

[0046] Δλ c =κ q Δq t +κ w Δw t

[0047] Where Δλ c The dimensionless model prediction compensation term; Δq t The current fuel calorific value q t Compared with the benchmark calorific value q r Deviation; Δw t The current fuel moisture content w t Compared with the baseline moisture content w r The deviation is dimensionless; q represents calorific value; t is used as a subscript here to indicate the current value; w represents moisture content;

[0048] The reference calorific value and reference moisture content are obtained based on the nominal values ​​of the reference fuel used in the boiler design, or by statistical averaging of the fuel characteristic data actually used in long-term operation.

[0049] κ q κ is the coefficient of influence on calorific value, with dimensions of mass / energy; w The coefficient κ represents the influence of moisture and is dimensionless. q With κ w These are preset parameters calibrated for specific boiler models and fuel types through offline regression analysis of historical combustion test data. The application of this formula lies in the fact that the state estimation module transforms the fuel characteristic data acquired by the Internet of Things into a quantified, forward-looking compensation signal Δλ. c This signal can predict changes in oxygen demand caused by fluctuations in fuel quality, thus providing crucial predictive information for subsequent state estimation and improving the system's adaptability and robustness to fuel disturbances.

[0050] This invention achieves proactive compensation for core disturbance sources. By setting up a data acquisition module, it not only acquires physical sensor data of the boiler, but more importantly, it proactively acquires fuel characteristic data characterizing the disturbance source, including fuel calorific value and fuel moisture content. Based on this, the state estimation module can calculate the model prediction compensation term Δλ. c =κ q Δq t +κ w Δw tThis formula, as stated in the instruction manual, is not derived from other formulas, but rather defined as a linear approximation model based on the principles of combustion stoichiometry. Its practical significance lies in relating fuel physical properties, such as calorific value q... t Moisture w t The change is directly quantified as the expected change in the excess air coefficient demand, Δλ. c This allows the system to pre-calculate the required compensation before inferior fuel enters the furnace and actually affects combustion, thus fundamentally solving the problem of delay caused by existing technologies that can only passively respond to results.

[0051] Example 3

[0052] The state estimation module is also used for:

[0053] Determine the dynamic fusion gain; sum the estimated instantaneous effective excess air coefficient of the previous moment with the model prediction compensation term to obtain the predicted value of the current moment; calculate the difference between the apparent excess air coefficient and the predicted value, and multiply it by the dynamic fusion gain to obtain the observation correction term; sum the observation correction term with the predicted value to generate the instantaneous effective excess air coefficient of the current moment.

[0054] Determine the dynamic fusion gain, including:

[0055] When the short-term variance of the apparent excess air coefficient is lower than a preset threshold, a larger preset value is assigned to the dynamic fusion gain.

[0056] When the short-term variance of the apparent excess air coefficient is not lower than a preset threshold, a preset small value is assigned to the dynamic fusion gain.

[0057] In this embodiment, the state estimation module obtains the model prediction compensation term Δλ. c Then, its core state estimation iterative expression is executed; the technical motivation of this algorithm is to integrate model prediction information and sensor measurement information to overcome the limitations of a single information source. Its principle draws on the idea of ​​state observers in modern control theory, aiming to obtain a more accurate, smoother, and more timely estimate of the system's internal state than any single information source; it is defined by the following iterative expression:

[0058] λ t =(λ t-1 +Δλ c )+K t (λ s -(λ t-1 +Δλ c ))

[0059] Where λ t λ represents the final estimated value of the instantaneous effective excess air coefficient at time t; λ represents the excess air system; λ t-1This is the estimated value from the previous time step; t-1 represents the previous time step; K represents the gain; t indicates that the gain is dynamic and related to time t; s is used as a subscript here, representing apparent or sensor-derived value; K t λ represents the dimensionless dynamic fusion gain. s Δλ is the apparent excess air coefficient calculated based on real-time flue gas oxygen concentration sensor data. c This is the model prediction compensation term calculated in the previous steps; the logical relationship of this expression is that the current best estimate λ t The value λ at the previous time step t-1 Based on this, through observation of the correction term K t (λ s -λ t-1 Feedback correction is performed, and a feedforward compensation term Δλ is added. c Predictive compensation is performed; short-term variance is based on the apparent excess air coefficient λ over a moving time window, e.g., the most recent 10 to 30 seconds, or a fixed number, e.g., the most recent 50 to 100 seconds. s The dynamic fusion gain K is calculated from the sample data. t The value selection logic is adaptive, and its determination method is as follows: when λ s The short-term variance is below a preset low threshold. When the sensor signal is stable, it indicates that the signal is K. t Assign a large preset value; when the variance exceeds a preset high threshold... When K indicates that the signal may contain noise or undergo drastic changes, it is considered to be K. t A small preset value is assigned; a larger value is typically set between 0.6 and 0.9 to ensure rapid tracking of the sensor measurement signal; while a smaller value is typically set between 0.1 and 0.4 to enhance the suppression of signal noise and the smoothness of the overall estimation; these specific values ​​can be ultimately determined during the system debugging phase by observing and optimizing the response characteristics under different operating conditions; threshold and All data were obtained through statistical analysis of historical stable operating conditions and typical disturbance operating conditions, ensuring reproducibility for those skilled in the art; this fusion mechanism ensures that subsequent control modules can make decisions based on a high-confidence state perception.

[0060] This invention achieves high-confidence real-time understanding of the internal state of the combustion process; the state estimation module employs a core state estimation iterative expression λ. t =(λ t-1 +Δλ c )+K t (λ s -(λ t-1 +Δλ cThe structure of this formula is a typical state observer; it is not a simple mathematical derivation, but rather an engineering application definition of control theory. It fuses three pieces of information: the state estimate λ from the previous time step. t-1 The inertia represents the system state; the observation correction term K t (λ s -λ t-1 ) Utilizing the measured value λ of a physical sensor s Feedback correction is applied to the estimate; while the feedforward compensation term Δλ c This incorporates predictions of future disturbances; by dynamically adjusting the fusion gain K t The system can intelligently weigh the degree of trust in sensor measurements; a direct technical effect is that the module outputs an estimated instantaneous effective excess air coefficient λ. t Compared to existing technologies that rely solely on noisy and delayed sensor signals, this technology offers higher accuracy and smoothness, providing a solid data foundation for subsequent precise control.

[0061] Example 4

[0062] The predictive control module is used to: subtract the preset target excess air coefficient from the instantaneous effective excess air coefficient to obtain the prediction deviation; calculate the time change rate of the instantaneous effective excess air coefficient; and calculate the feedforward control adjustment amount based on the prediction deviation, the time change rate, the preset proportional feedforward gain, and the preset differential feedforward gain. In this embodiment, the predictive control module receives the high-precision estimated value λ output by the state estimation module. t and its rate of change over time The core task is to generate a feedforward control adjustment variable Δu. f The technical motivation behind this feedforward control law lies in overcoming the inherent lag of traditional feedback control. Its underlying logic is that the control action should be proportional to the predicted deviation and its changing trend, thus intervening at the initial stage of deviation formation. It is defined by a proportional-derivative feedforward controller.

[0063]

[0064] Where Δu f (t) represents the dimensionless feedforward control adjustment; Δu represents the control adjustment; f is a subscript here representing feedforward; p is a subscript here representing proportional; λ r The optimal excess air coefficient target setpoint determined based on the current boiler load; r is used as a subscript here, representing the target or setpoint; λ t The high-precision instantaneous effective excess air coefficient estimate output by the state estimation module; k d Differential feedforward gain, its dimension is time; d is used here as a subscript to represent the derivative; λ rThe target setpoint for the optimal excess air coefficient is determined based on the current boiler load; (λ) r -λ t ) represents the prediction bias; The rate of change of the estimated value is expressed as the reciprocal of time; k p k is the dimensionless proportional feedforward gain. d The gain is a differential feedforward gain, with the dimension being time, which ensures... Dimensionless transformation of the term; gain parameter k p With k d The value of is tuned through step response experiments on a real system; the application effect of this algorithm is to endow the system with strong dynamic disturbance suppression capability; the proportional term responds to the current deviation, while the differential term predicts the trend of deviation change, so that when λ t When rapidly approaching the undesirable region, even with a small current deviation, significant early intervention can be achieved, effectively preventing overshoot and oscillation in the combustion state. This invention realizes a transformation in control mode from passive correction to active intervention. The predictive control module is based on a high-precision state estimation value λ. t Through feedforward control law This is used to calculate the feedforward control adjustment; this formula is also a definition of a control algorithm; its significance lies in the fact that its control effect is not only proportional to the current deviation (λ) r -λ t It also includes the trend of deviation changes. Predictions; for example, when the boiler load suddenly increases, even if λ t It has not deviated significantly from the target value λ r However, its rapid downward trend will be captured by the differential term, thereby generating a powerful and proactive control effect that increases the airflow. This initiative changes the situation where existing technologies must wait for error accumulation before responding, greatly improving the dynamic response speed and stability of the system.

[0065] Example 5

[0066] The instruction fusion module is used for:

[0067] Determine the feedforward weighting coefficient; multiply the feedforward control adjustment by the feedforward weighting coefficient to obtain the weighted feedforward quantity;

[0068] The weighted feedback quantity is obtained by multiplying the corrective control command by the difference between the feedforward weighting coefficient and the weighted feedforward quantity. The weighted feedforward quantity, the weighted feedback quantity, and the basic control signal determined according to the boiler load are summed to generate the final control command.

[0069] Determine the feedforward weighting coefficients, including:

[0070] When the absolute value of the model's predicted compensation term or the absolute value of the time-varying rate of change of the instantaneous effective excess air coefficient exceeds the preset disturbance threshold, the feedforward weight coefficient is increased.

[0071] When the absolute value of the model's predicted compensation term and the absolute value of the time-varying rate of change of the instantaneous effective excess air coefficient do not exceed the preset disturbance threshold, the feedforward weight coefficient is reduced.

[0072] In this embodiment, the instruction fusion module is responsible for combining the advantages of feedforward control and feedback control. The technical motivation for this fusion strategy is to form a predictive-corrective composite control structure, utilizing the fast response of feedforward control to handle measurable disturbances, while utilizing the error correction capability of feedback control to handle unmodeled disturbances and steady-state errors. This structure is achieved through the following dynamic weighted fusion formula:

[0073] u t =u b +β·Δu f +(1-β)·u p

[0074] Where u t This refers to the final control signal sent to the field actuator; u represents the control signal or command; t is a subscript here, possibly indicating final; b is a subscript here, indicating basic; p is a subscript here, indicating a corrective command from the feedback loop; u b The basic control signal, Δu, is determined by a preset lookup table method based on the current boiler load; f This is the adjustment amount for feedforward control; u p The feedback control output comes from a standard PID controller; β is a dimensionless dynamic feedforward weighting coefficient; the logic for determining β is event-driven adaptive adjustment: when the system detects a significant disturbance, i.e., the absolute value of the model prediction compensation term |Δλ... c |or the absolute value of the estimated rate of change When the disturbance exceeds a preset threshold, the value of β is increased, making the control response primarily feedforward; under steady-state conditions, all the above indicators are below the threshold, and the value of β is decreased to enhance the fine adjustment effect of PID feedback control.

[0075] The increased feedforward weighting coefficient is typically set in the range of 0.7-0.9 to prioritize the rapid response capability of feedforward control when disturbances occur. Under steady-state conditions, the decreased value is typically set in the range of 0.2-0.4 to focus on the fine adjustment of feedback control and eliminate steady-state errors. These weighting coefficient values ​​can be finally calibrated during system commissioning based on the suppression effect on typical disturbances and the smoothness of steady-state operation.

[0076] The disturbance threshold here is set based on statistical analysis of historical disturbance data, ensuring the reliability and reproducibility of the judgment; this fusion strategy realizes smooth switching of control modes, enabling the system to exert the speed of predictive control and the accuracy of feedback control under any operating condition, achieving the best global control effect.

[0077] This invention achieves an optimal balance between speed and precision in control; the instruction fusion module controls the final instruction u. t =u b +β·Δu f +(1-β)·u p The system achieves dynamic coordination of both feedforward and feedback control strategies; the adaptive adjustment logic of the weight coefficient β is key; when a large disturbance is detected, the system increases the β value, relying on the speed of feedforward control to quickly stabilize the situation; under stable conditions, the β value is reduced to leverage the accuracy of feedback control to eliminate steady-state error; this design overcomes the limitations of a single control mode in the prior art, enabling the system of this invention to exhibit optimal control performance under any operating condition.

[0078] In summary, this invention transforms passive, error-based feedback control into active, prediction-based feedforward-feedback composite control by constructing a complete logical closed loop from data acquisition, state estimation, predictive control to command fusion. The resulting benefits are a significant improvement in the combustion efficiency of industrial boilers and a reduction in fuel consumption. Simultaneously, because the combustion process is more precisely controlled at its optimal state, the generation of incomplete combustion products such as carbon monoxide and nitrogen oxides caused by excess air is effectively suppressed, resulting in significant environmental benefits. Furthermore, the overall operational robustness and adaptability to external disturbances of the system far exceed those of existing technologies.

[0079] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A real-time monitoring and energy efficiency optimization system for the combustion status of industrial boilers based on the Internet of Things, characterized in that, include: The data acquisition module is used to acquire physical sensor data and fuel characteristic data of the boiler. The physical sensor data includes flue gas oxygen concentration, and the fuel characteristic data includes fuel calorific value and fuel moisture content. The state estimation module is used for: The calorific value deviation is determined based on the difference between the fuel calorific value and the preset reference calorific value. The moisture deviation is determined based on the difference between the fuel moisture content and the preset reference moisture content; Multiply the calorific value deviation by a preset calorific value influence coefficient to obtain the first product; Multiply the moisture deviation by a preset moisture influence coefficient to obtain a second product; Sum the first product and the second product to calculate the model prediction compensation term; The apparent excess air coefficient is determined based on the flue gas oxygen concentration. Determine the dynamic fusion gain; The estimated value of the instantaneous effective excess air coefficient at the previous moment is summed with the model prediction compensation term to obtain the predicted value at the current moment. Calculate the difference between the apparent excess air coefficient and the predicted value at the current time, and multiply the difference by the dynamic fusion gain to obtain the observation correction term; The observed correction term is summed with the predicted value to generate the instantaneous effective excess air coefficient at the current moment; Determining the dynamic fusion gain includes: When the short-term variance of the apparent excess air coefficient is lower than a preset threshold, a preset large value is assigned to the dynamic fusion gain. When the short-term variance of the apparent excess air coefficient is not lower than the preset threshold, a preset small value is assigned to the dynamic fusion gain. Predictive control module, used for: The prediction deviation is obtained by subtracting the preset target excess air coefficient from the instantaneous effective excess air coefficient. Calculate the rate of change of the instantaneous effective excess air coefficient over time; Based on the prediction deviation, the time change rate, the preset proportional feedforward gain, and the preset differential feedforward gain, the feedforward control adjustment amount is calculated. The instruction fusion module is used for: Determine the feedforward weighting coefficients; Multiply the feedforward control adjustment by the feedforward weighting coefficient to obtain the weighted feedforward amount; The weighted feedback quantity is obtained by multiplying the corrective control command generated by the feedback control loop by the difference obtained by subtracting the feedforward weight coefficient. The weighted feedforward quantity, the weighted feedback quantity, and the basic control signal determined according to the boiler load are summed to generate the final control command; When the absolute value of the model prediction compensation term or the absolute value of the time change rate of the instantaneous effective excess air coefficient exceeds the preset disturbance threshold, the feedforward weight coefficient is increased. When the absolute value of the model prediction compensation term and the absolute value of the time change rate of the instantaneous effective excess air coefficient do not exceed the preset disturbance threshold, the feedforward weight coefficient is reduced to regulate the combustion process of the boiler.

Images