Boiler energy efficiency intelligent management method and device and medium

Abstract

The invention discloses an intelligent boiler energy efficiency management method and device and a medium, and relates to the technical field of boiler energy efficiency management.The intelligent boiler energy efficiency management method includes the steps that firstly, hearth flame images, smoke components and water-cooled wall temperature difference data are collected in real time, and a dynamic behavior model of boiler operation is established; the flame stability, the denitration catalyst activity interval and the water wall temperature difference development trend under the low-load working condition are predicted, and the prediction capacity for the complex working condition is improved; then, on the basis of the dynamic behavior model, a comprehensive target of combustion efficiency maximization, pollutant emission minimization and equipment service life prolonging is set through a particle swarm optimization algorithm, and an optimal adjustment strategy of the fuel injection rate, air flow distribution and the water treatment agent putting amount is generated; besides, an adjustment strategy is decoded and transmitted in real time, an intelligent controller directly adjusts air distribution of a combustor, denitration ammonia spraying and operation parameters of a circulating pump, closed-loop optimization is carried out in combination with feedback of a sensor, and the intelligent level and stability of boiler operation are effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of boiler energy efficiency management, and particularly to a method, device and medium for intelligent management of boiler energy efficiency. Background Art

[0002] When coal-fired boilers operate at low load for peak shaving, they face problems such as poor combustion stability, reduced denitration efficiency, overheating of water-cooled walls, and equipment wear. These problems seriously affect the reliability and economy of operation; low-load operation leads to a reduction in fuel supply, a significant decrease in the stability of the furnace flame, a deviation of combustion characteristics from design parameters, and is prone to flameout accidents; at the same time, due to the decrease in flue gas temperature, the activity of the denitration catalyst weakens, and the denitration efficiency decreases significantly, and even fails to meet the environmental protection emission standards.

[0003] Regarding the problem of poor combustion stability, most traditional boiler management schemes stabilize the flame by adjusting the ratio of fuel to air or increasing the injection amount of auxiliary fuel oil. Although this method is effective in the short term, it is difficult to quickly respond to complex and changeable working conditions; to improve the problem of decreased denitration efficiency, some schemes increase the ammonia injection amount or use preheating equipment to increase the catalyst temperature, but these measures will increase energy consumption and operating costs, and may cause new problems such as ammonia escape; therefore, there is an urgent need for an intelligent management method for boiler energy efficiency to solve such problems. Summary of the Invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] The present invention provides a method, device and medium for intelligent management of boiler energy efficiency to solve the problems that the combustion adjustment and denitration optimization of traditional boiler energy efficiency intelligent management schemes rely on static rules, lack real-time intelligent adjustment capabilities, and are difficult to meet the requirements of frequent changes in working conditions during low-load operation; while auxiliary measures such as oil injection for combustion assistance and preheating can improve stability and denitration efficiency, but increase operating costs and may cause secondary pollution problems.

[0006] To solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, an embodiment of the present invention provides a method for intelligent management of boiler energy efficiency, which includes,

[0008] Step S1, deploying a sensor network at the boiler operation node to collect energy efficiency data, including furnace flame images, flue gas temperature and composition, water-cooled wall temperature distribution, and water quality data;

[0009] Step S2, using a reinforcement learning algorithm to analyze the energy efficiency data and establish a dynamic behavior model of boiler operation;

[0010] Step S3, set the optimization objectives and constraints based on the dynamic behavior model, including combustion efficiency, pollutant emissions, and equipment life, and use the particle swarm optimization algorithm to generate adjustment strategies, including fuel injection rate, air flow distribution, and water treatment agent dosage;

[0011] Step S4, transfer the adjustment strategy to the burner, denitration device, and water treatment equipment, and adjust the air distribution strategy, ammonia injection amount, and circulating pump operation parameters of the burner in real time.

[0012] As a preferred solution of the intelligent management method for boiler energy efficiency described in the present invention, wherein: the dynamic behavior model predicts the flame stability at low load, the active interval of the denitration catalyst, and the development trend of the water wall temperature difference.

[0013] As a preferred solution of the intelligent management method for boiler energy efficiency described in the present invention, wherein: the step of using the reinforcement learning algorithm to analyze the energy efficiency data and establish the dynamic behavior model of boiler operation is as follows:

[0014] Construct a reinforcement learning framework, including:

[0015] State space Is a set of real-time operation parameters of the boiler, defined as:

[0016] Wherein, T f Is the furnace temperature, Is the oxygen content in the flue gas, Is the concentration of nitrogen oxides in the flue gas, ΔT w Is the water wall temperature difference, L b Is the boiler load level;

[0017] Action space Is a set of controllable parameter adjustments, defined as:

[0018] Wherein, Is the fuel injection rate, Is the air flow rate, Is the water treatment agent dosage, Is the ammonia injection rate;

[0019] Reward function R, integrating energy efficiency, emissions, and equipment status optimization objectives, is defined as:

[0020] Wherein, η is the boiler combustion efficiency, P penalty Is the equipment safety penalty factor, w 1 , w 2 , w 3 , w 4is a weight parameter that affects the state through actions and feedbacks the effect of each adjustment according to the reward function.

[0021] As a preferred solution of the intelligent management method for boiler energy efficiency described in the present invention, wherein: the construction steps of the dynamic behavior model are as follows.

[0022] Based on the energy efficiency data in step S1, use the time series modeling method to extract features:

[0023] The input is the sensor data X, X = {x t |t = 1, 2,..., T}, where x t is the input data at the current time step;

[0024] Use the long short-term memory network LSTM to extract time series features, and the extraction formula is:

[0025] h t = σ(W h x t + U h h t-1 + b h ), where h t is the hidden state at time t, h t-1 represents the hidden state at time t - 1, W h , U h , b h are the input weight matrix, recurrent weight matrix and bias vector of the LSTM respectively, and σ is the activation function;

[0026] Use reinforcement learning to train the dynamic behavior model, and the goal is to maximize the cumulative reward. The training formula is:

[0027]

[0028] Among them, J(π) is the long-term return function of the policy π, E[·] is the expectation function, γ is the discount factor, and R t is the immediate feedback signal;

[0029] The dynamic behavior model outputs the predicted values of the key boiler performance. The model formula is:

[0030]

[0031] Among them, is the predicted furnace temperature, is the predicted nitrogen oxide emission concentration, is the predicted water wall temperature difference, represents the output of the dynamic behavior model.

[0032] As a preferred solution of the intelligent management method for boiler energy efficiency according to the present invention, in the step S2, the characteristics under low load are predicted by using a dynamic behavior model, and the steps are as follows:

[0033] The flame stability is predicted by using a dynamic behavior model, and the prediction formula is:

[0034]

[0035] Wherein, is the critical temperature of flame extinction, is the peak temperature of the flame, and S f > 0.8 represents the stability index;

[0036] The active interval of the denitration catalyst is predicted by using a dynamic behavior model, and the prediction formula is:

[0037]

[0038] Wherein, [T min , T max is the catalyst design temperature range, δ is the temperature fluctuation range, and A cat is the active interval of the denitration catalyst;

[0039] The development trend of the water wall temperature difference is predicted by using a dynamic behavior model, and the prediction formula is:

[0040] Wherein, is the predicted value of the water wall temperature difference.

[0041] As a preferred solution of the intelligent management method for boiler energy efficiency according to the present invention, the steps of setting optimization objectives and constraint conditions based on the dynamic behavior model are as follows:

[0042] Set optimization objectives, including combustion efficiency, pollutant emission control, and equipment life extension. The output indexes of the dynamic behavior model are defined as:

[0043] The combustion efficiency objective function is:

[0044] η = Q useful / Q input ,

[0045] Wherein, Q useful is the effective heat output, Q input is the fuel input heat, and the strengthening objective is to maximize η;

[0046] The pollutant emission objective function is:

[0047] Minimize pollutant emissions, for nitrogen oxides and sulfur dioxide Emission constraints: Among them, C env is the defined emission standard;

[0048] The objective function for equipment life extension is:

[0049] With the temperature difference ΔT of the water wall w as the constraint: ΔT w ≤ΔT safe , among which, ΔT safe is the maximum allowable temperature difference,

[0050] The comprehensive optimization objective J is:

[0051]

[0052] Among them, w 1 , w 2 , w 3 are the weight factors of combustion efficiency, emissions, and equipment life.

[0053] As a preferred solution of the intelligent management method for boiler energy efficiency described in the present invention, among them: the step of setting the optimization objective and constraint conditions further includes,

[0054] Defining constraint conditions based on the predicted values of the dynamic behavior model , including:

[0055] Flame stability constraint:

[0056]

[0057] Denitration catalyst activity constraint:

[0058]

[0059] Safety constraint:

[0060] ΔT w ≤ΔT safe ,

[0061] The step of generating an adjustment strategy by using the particle swarm optimization algorithm is,

[0062] Defining that a particle represents a set of boiler operation control parameters p i ,

[0063]

[0064] Among them, is the fuel injection rate of the i-th particle, is the air flow distribution, is the dosage of water treatment agent, is the ammonia injection rate,

[0065] At each iteration, update the particle position according to the velocity. The particle position update formula is:

[0066] where, is the parameter position of particle i at the t-th iteration, is the parameter position of particle i at the (t + 1)-th iteration, is the particle velocity;

[0067] The particle velocity update formula is:

[0068]

[0069] where, is the particle velocity at the (t + 1)-th iteration, ω is the inertia weight, c 1 , c 2 is the learning factor, r 1 , r 2 is a random number, p best is the historical best position of particle i, g best is the global best position;

[0070] Calculate the fitness function J for each particle to judge its optimization effect. The judgment formula is:

[0071] where, J i represents the fitness value of particle i, w 1 η i is the combustion efficiency contribution value of particle i, and the weight w 1 controls the priority of efficiency, represents the emission index penalty value of the particle, represents the equipment life risk penalty value of the particle;

[0072] When the number of iterations reaches the upper limit or the improvement amplitude of the global optimal solution is lower than the threshold, output the optimal adjustment strategy.

[0073] As a preferred solution of the intelligent management method for boiler energy efficiency described in the present invention, wherein: the step of transmitting the adjustment strategy to the burner, the denitration device and the water treatment equipment and adjusting the air distribution strategy, the ammonia injection amount and the circulating pump operation parameters of the burner in real time is,

[0074] Decode the adjustment strategy. Let the optimal parameter output by the adjustment strategy be P opt :

[0075]

[0076] Transfer Popt Decoded into device control instructions, including:

[0077] Controlling the fuel injection rate and air distribution of the burner,

[0078] Adjusting the ammonia injection amount of the denitration device,

[0079] Controlling the dosing amount of the circulating pump water treatment agent of the water treatment equipment;

[0080] Calculating the actual device control parameters,

[0081] The fuel - to - air ratio of the burner is:

[0082] The ammonia injection concentration is adjusted to:

[0083]

[0084] Among them, C NH3 is the ammonia injection concentration for denitration, is the flue gas flow rate, measured by a flow meter,

[0085] The concentration of the water treatment agent is:

[0086]

[0087] Among them, C water is the concentration of the water treatment agent, is the circulating water flow rate;

[0088] The steps of the real - time adjustment of the air distribution strategy, ammonia injection amount and circulating pump operation parameters of the burner further include: the real - time controller should adjust the strategy,

[0089] Using the burner controller to control the fuel injection rate and the air flow rate to perform air distribution optimization,

[0090] Using the denitration controller to adjust the ammonia injection rate to maintain the active range of the denitration catalyst,

[0091] Using the water treatment controller to control the dosing amount of the circulating pump water treatment agent to keep the water quality stable;

[0092] Each controller receives the sensor feedback data and performs real - time correction according to the adjusted operation parameters. The feedback mechanism formula is: u k = u k-1 + K e ·(y set - y k), where u k is the control parameter after the k-th adjustment, y set is the target value, y k is the current feedback value, K e is the feedback gain.

[0093] In a second aspect, an embodiment of the present invention provides a computer device, including a memory and a processor. The memory stores a computer program, where: when the computer program is executed by the processor, any step of the intelligent management method for boiler energy efficiency described in the first aspect of the present invention is implemented.

[0094] In a third aspect, an embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored, where: when the computer program is executed by the processor, any step of the intelligent management method for boiler energy efficiency described in the first aspect of the present invention is implemented.

[0095] The beneficial effects of the present invention are as follows: First, the present invention collects the furnace flame image, flue gas composition, and water wall temperature difference data in real time, and combines reinforcement learning and long short-term memory network (LSTM) technology to establish a dynamic behavior model of boiler operation, predicting the flame stability, the active interval of the denitration catalyst, and the development trend of the water wall temperature difference under low load conditions, improving the prediction ability for complex working conditions. Then, based on the dynamic behavior model, through the particle swarm optimization algorithm, a comprehensive goal of maximizing combustion efficiency, minimizing pollutant emissions, and extending equipment life is set, generating an optimal adjustment strategy for fuel injection rate, air flow distribution, and water treatment agent dosage. In addition, the adjustment strategy is decoded and transmitted in real time, and the intelligent controller directly adjusts the burner air distribution, denitration ammonia injection, and circulating pump operation parameters, and combines sensor feedback for closed-loop optimization, effectively improving the intelligent level and stability of boiler operation. It not only improves the combustion and denitration efficiency under low load operation, but also extends the equipment service life, reduces the operation cost and the frequency of unplanned shutdowns. Description of the Drawings

[0096] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0097] Figure 1 It is a flowchart of an intelligent management method for boiler energy efficiency in Embodiment 1. Detailed Embodiments

[0098] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following will make a detailed description of the specific embodiments of the present invention with reference to the drawings in the specification.

[0099] In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention. However, the present invention may be practiced in other ways than those specifically described herein, and those skilled in the art can make similar extensions without departing from the spirit of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0100] Secondly, as used herein, "one embodiment" or "an embodiment" refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The appearances of "in one embodiment" in different places in this specification do not all refer to the same embodiment, nor are they separate or alternative embodiments that exclude each other.

[0101] Embodiment 1, referring to Figure 1 , this embodiment provides an intelligent management method for boiler energy efficiency, including the following steps:

[0102] Step S1, deploy a sensor network at the boiler operation node to collect energy efficiency data, including furnace flame images, flue gas temperature and composition, water wall temperature distribution, and water quality data;

[0103] Step S2, analyze the energy efficiency data using a reinforcement learning algorithm and establish a dynamic behavior model of the boiler operation;

[0104] The dynamic behavior model predicts the flame stability at low load, the active interval of the denitration catalyst, and the development trend of the water wall temperature difference;

[0105] The steps of analyzing the energy efficiency data using a reinforcement learning algorithm and establishing a dynamic behavior model of the boiler operation are as follows:

[0106] Construct a reinforcement learning framework, including:

[0107] State space Is a set of real-time operation parameters of the boiler, defined as:

[0108] Wherein, T f Is the furnace temperature, Is the oxygen content in the flue gas, Is the concentration of nitrogen oxides in the flue gas, ΔT w Is the water wall temperature difference, L b Is the boiler load level;

[0109] Action space Is a set of controllable parameter adjustments, defined as:

[0110] Wherein, Is the fuel injection rate, is the air flow rate, is the dosage of water treatment agent, is the ammonia injection rate;

[0111] The reward function R, which comprehensively optimizes the goals of energy efficiency, emissions, and equipment status, is defined as:

[0112] where η is the boiler combustion efficiency, P penalty is the equipment safety penalty factor, w 1 , w 2 , w 3 , w 4 are the weight parameters, which affect the state through actions and feedback the effect of each adjustment according to the reward function;

[0113] The steps for constructing the dynamic behavior model are as follows:

[0114] Based on the energy efficiency data in step S1, use the time series modeling method to extract features:

[0115] The input is the sensor data X, X = {x t |t = 1, 2,..., T}, where x t is the input data at the current time step;

[0116] Use the long short-term memory network LSTM to extract time series features. The extraction formula is:

[0117] h t = σ(W h x t + U h h t-1 + b h ), where h t is the hidden state at time t, h t-1 represents the hidden state at time t - 1, W h , U h , b h are the input weight matrix, recurrent weight matrix, and bias vector of the LSTM respectively, and σ is the activation function;

[0118] Use reinforcement learning to train the dynamic behavior model. The goal is to maximize the cumulative reward. The training formula is:

[0119]

[0120] where J(π) is the long-term return function of the policy π, E[·] is the expectation function, γ is the discount factor, and R t is the immediate feedback signal;

[0121] The dynamic behavior model outputs the predicted values of the key boiler performance. The model formula is:

[0122]

[0123] Among them, is the predicted furnace temperature, is the predicted concentration of nitrogen oxides emissions, is the predicted temperature difference of the water wall, represents the output of the dynamic behavior model;

[0124] In step S2, the dynamic behavior model is used to predict the characteristics under low load. The steps are as follows:

[0125] The dynamic behavior model is used to predict the flame stability. The prediction formula is:

[0126]

[0127] Among them, is the critical temperature of flame extinction, is the peak flame temperature, S f > 0.8 represents the stability index;

[0128] The dynamic behavior model is used to predict the active interval of the denitration catalyst. The prediction formula is:

[0129]

[0130] Among them, [T min , T max is the designed temperature range of the catalyst, δ is the temperature fluctuation range, and A cat is the active interval of the denitration catalyst;

[0131] The dynamic behavior model is used to predict the development trend of the water wall temperature difference. The prediction formula is:

[0132] Among them, is the predicted value of the water wall temperature difference;

[0133] Specifically, a dynamic behavior model of boiler operation is established through reinforcement learning and time series analysis to predict key performance indicators;

[0134] In step S3, based on the dynamic behavior model, optimization objectives and constraint conditions are set, including combustion efficiency, pollutant emissions, and equipment life. The particle swarm optimization algorithm is used to generate adjustment strategies, including fuel injection rate, air flow distribution, and water treatment agent dosage;

[0135] The steps for setting optimization objectives and constraint conditions based on the dynamic behavior model are as follows:

[0136] Set optimization objectives, including combustion efficiency, pollutant emission control, and equipment life extension. The output indicators of the dynamic behavior model are defined as:

[0137] The combustion efficiency objective function is:

[0138] η = Q useful / Q input ,

[0139] where Q useful is the effective heat output, Q input is the fuel input heat, and the strengthening objective is to maximize η;

[0140] The pollutant emission objective function is:

[0141] Minimize pollutant emissions, with emission constraints on nitrogen oxides and sulfur dioxide : where C env is the defined emission standard;

[0142] The equipment life extension objective function is:

[0143] With the temperature difference ΔT w of the water wall as the constraint: ΔT w ≤ΔT safe , where ΔT safe is the maximum allowable temperature difference,

[0144] The comprehensive optimization objective J is:

[0145]

[0146] where w 1 , w 2 , w 3 are the weight factors for combustion efficiency, emissions, and equipment life respectively;

[0147] The steps to set the optimization objectives and constraints also include,

[0148] Defining the constraints based on the predicted values of the dynamic behavior model , including:

[0149] Flame stability constraint:

[0150]

[0151] De-NOx catalyst activity constraint:

[0152]

[0153] Safety constraint:

[0154] ΔT w ≤ΔT safe ,

[0155] The steps to generate the adjustment strategy using the particle swarm optimization algorithm are as follows:

[0156] Define a particle to represent a set of boiler operation control parameters p i ,

[0157]

[0158] where, is the fuel injection rate of the i-th particle, is the air flow distribution, is the dosage of water treatment agent, is the ammonia injection rate,

[0159] In each iteration, update the particle position according to the velocity. The particle position update formula is:

[0160] where, is the parameter position of particle i in the t-th iteration, is the parameter position of particle i in the (t + 1)-th iteration, is the particle velocity;

[0161] The particle velocity update formula is:

[0162]

[0163] where, is the particle velocity in the (t + 1)-th iteration, ω is the inertia weight, c 1 , c 2 is the learning factor, r 1 , r 2 is a random number, p best is the historical optimal position of particle i, g best is the global optimal position;

[0164] Calculate the fitness function J for each particle to judge its optimization effect. The judgment formula is:

[0165] where, J i represents the fitness value of particle i, w 1 η i is the contribution value of the combustion efficiency of particle i, and the weight w 1 controls the priority of the control efficiency, represents the penalty value of the emission index of the particle, represents the penalty value of the equipment life risk of the particle;

[0166] When the number of iterations reaches the upper limit or the improvement amplitude of the global optimal solution is lower than the threshold, output the optimal adjustment strategy;

[0167] Specifically, under the constraint conditions of the dynamic behavior model, the particle swarm optimization algorithm generates the optimal adjustment strategies for the fuel injection rate, air flow distribution, and water treatment agent dosage, achieving the comprehensive goals of maximizing combustion efficiency, minimizing emissions, and extending equipment life.

[0168] Step S4, transfer the adjustment strategy to the burner, denitration device, and water treatment equipment, and adjust the air distribution strategy, ammonia injection amount, and circulating pump operation parameters of the burner in real time;

[0169] The step of transferring the adjustment strategy to the burner, denitration device, and water treatment equipment and adjusting the air distribution strategy, ammonia injection amount, and circulating pump operation parameters of the burner in real time is as follows:

[0170] Perform adjustment strategy decoding, and set the optimal parameter output by the adjustment strategy as P opt :

[0171]

[0172] Transfer P opt Decode it into equipment control instructions, including:

[0173] Control the fuel injection rate and air distribution of the burner,

[0174] Adjust the ammonia injection amount of the denitration device,

[0175] Control the water treatment agent dosage of the circulating pump of the water treatment equipment;

[0176] Calculate the actual equipment control parameters,

[0177] The fuel-to-air ratio of the burner is: Among them, φ is the fuel-to-air equivalence ratio,

[0178] The ammonia injection concentration is adjusted to:

[0179]

[0180] Among them, is the denitration ammonia injection concentration, is the flue gas flow rate, measured by a flow meter,

[0181] The water treatment agent concentration is:

[0182]

[0183] Among them, C water is the water treatment agent concentration, is the circulating water flow rate;

[0184] The steps of real-time adjusting the air distribution strategy, ammonia injection amount and circulating pump operation parameters of the burner further include: the real-time controller should adjust the strategy,

[0185] using the burner controller to control the fuel injection rate and the air flow rate performing air distribution optimization,

[0186] using the denitration controller to adjust the ammonia injection rate to maintain the active range of the denitration catalyst,

[0187] using the water treatment controller to control the dosage of water treatment agent for the circulating pump to keep the water quality stable;

[0188] Each controller receives the sensor feedback data and performs real-time correction according to the adjusted operation parameters. The feedback mechanism formula is: u k = u k-1 + K e ·(y set - y k ), where u k is the control parameter after the k-th adjustment, y set is the target value, y k is the current feedback value, and K e is the feedback gain;

[0189] Specifically, through adjusting the strategy decoding and mapping and applying the real-time controller, the adjustment strategy is transmitted and applied to the boiler burner, denitration device and water treatment equipment to realize the real-time optimization and dynamic regulation of combustion, emissions and equipment protection.

[0190] This embodiment also provides a computer device, applicable to the situation of a boiler energy efficiency intelligent management method, device and medium, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement a boiler energy efficiency intelligent management method, device and medium as proposed in the above embodiment.

[0191] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a carrier network, NFC (Near Field Communication), or other technologies. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, a trackball, or a touchpad provided on the outer shell of the computer device, or an external keyboard, touchpad, or mouse, etc.

[0192] This embodiment also provides a storage medium, on which a computer program is stored. When the program is executed by a processor, it implements a boiler energy efficiency intelligent management method, device, and medium as proposed in the above embodiment; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, abbreviated as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, abbreviated as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, abbreviated as EPROM), programmable read-only memory (Programmable Red-Only Memory, abbreviated as PROM), read-only memory (Read-Only Memory, abbreviated as ROM), magnetic memory, flash memory, a magnetic disk, or an optical disc.

[0193] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered within the scope of the claims of the present invention.

Claims

1. A boiler energy efficiency intelligent management method, characterized by: include, Step S1, deploying a sensor network at the boiler operation node to collect energy efficiency data, including furnace flame image, flue gas temperature and composition, water wall temperature distribution and water quality data; Step S2, using a reinforcement learning algorithm to analyze the energy efficiency data and establish a dynamic behavior model of boiler operation; Step S3, setting optimization objectives and constraints based on the dynamic behavior model, including combustion efficiency, pollutant emissions and equipment life, and using a particle swarm optimization algorithm to generate an adjustment strategy, including fuel injection rate, air flow distribution and water treatment agent dosage; Step S4, transferring the adjustment strategy to the burner, the denitrification device and the water treatment equipment, and adjusting the air distribution strategy, the ammonia injection amount and the circulation pump operating parameters of the burner in real time.

2. A boiler energy efficiency intelligent management method according to claim 1, characterized in that: The dynamic behavior model predicts the flame stability under low load, the activity range of the denitration catalyst and the development trend of the water-cooled wall temperature difference.

3. A boiler energy efficiency intelligent management method according to claim 2, characterized in that: The steps of using the reinforcement learning algorithm to analyze the energy efficiency data and establish a dynamic behavior model of boiler operation are: Build a reinforcement learning framework, including: State Space is a set of real-time operating parameters of the boiler, defined as: Among them, T f is the furnace temperature, is the oxygen content in the flue gas, is the concentration of nitrogen oxides in flue gas, ΔT w is the water wall temperature difference, L b is the boiler load level; Action Space is the set of controllable parameter adjustments, defined as: in, is the fuel injection rate, is the air flow rate, is the amount of water treatment agent added, is the ammonia injection rate; The reward function R, which integrates energy efficiency, emissions and equipment status optimization objectives, is defined as: Where η is the boiler combustion efficiency, P penalty is the device security penalty factor, w1,w2,w3,w4 are weight parameters.

4. A boiler energy efficiency intelligent management method according to claim 3, characterized in that: The steps of constructing the dynamic behavior model are: Based on the energy efficiency data in step S1, the time series modeling method is used to extract features: The input is sensor data X, X = {x t |t=1,2,...,T}, where x t is the input data of the current time step; The long short-term memory network LSTM is used to extract time series features. The extraction formula is: h t =σ(W h x t +U h h t-1 +b h ), where h t is the hidden state at time t, h t-1 represents the hidden state at time t-1, W h ,U h ,b h are the input weight matrix, cyclic weight matrix and bias vector of LSTM respectively, and σ is the activation function; The dynamic behavior model is trained using reinforcement learning. The goal is to maximize the cumulative reward. The training formula is: Among them, J(π) is the long-term return function of strategy π, E[·] is the expectation function, γ is the discount factor, and R t It is an immediate feedback signal; The dynamic behavior model outputs the predicted value of the boiler's key performance. The model formula is: in, is the predicted furnace temperature, is the predicted NOx emission concentration, is the predicted water wall temperature difference, Represents the output of the dynamic behavior model.

5. A boiler energy efficiency intelligent management method according to claim 4, characterized in that: In step S2, the characteristics under low load are predicted using a dynamic behavior model, the steps are: The dynamic behavior model is used to predict flame stability, and the prediction formula is: in, is the critical temperature for flame extinction, is the flame peak temperature, S f >0.8 indicates stability index; The dynamic behavior model is used to predict the activity range of the denitration catalyst. The prediction formula is: Among them, [T min ,T max ] is the catalyst design temperature range, δ is the temperature fluctuation range, A cat is the activity range of the denitration catalyst; The dynamic behavior model is used to predict the development trend of the water-cooled wall temperature difference. The prediction formula is: in, is the predicted value of water wall temperature difference.

6. A boiler energy efficiency intelligent management method according to claim 5, characterized in that: The steps of setting optimization objectives and constraints based on the dynamic behavior model are: The optimization goals are set, including combustion efficiency, pollutant emission control and equipment life extension. The output indicators of the dynamic behavior model are defined as: The combustion efficiency objective function is: η=Q useful / Q input , Among them, Q useful is the effective heat output, Q input Input heat to the fuel, the strengthening goal is to maximize η; The pollutant emission objective function is: Minimize pollutant emissions, nitrogen oxides and sulfur dioxide Emission constraints: Among them, C env To define emission standards; The equipment life extension objective function is: The water wall temperature difference ΔT w Constraint: ΔT w ≤ΔT safe , where ΔT safe is the maximum allowable temperature difference, The comprehensive optimization objective J is: Among them, w1, w2, w3 are weight factors of combustion efficiency, emissions and equipment life.

7. A boiler energy efficiency intelligent management method according to claim 6, characterized in that: The step of setting the optimization goal and constraints also includes: Predicted values ​​based on dynamic behavior models Define constraints, including: Flame stability constraints: DeNOx catalyst activity constraints: Security constraints: The step of using the particle swarm optimization algorithm to generate the adjustment strategy is: Define a particle to represent a set of boiler operation control parameters p i , in, is the fuel injection rate of the ith particle, is the air flow distribution, is the amount of water treatment agent added, is the ammonia injection rate, In each iteration, the particle position is updated according to the velocity, and the particle position update formula is: in, is the parameter position of particle i at the tth iteration, is the parameter position of particle i at the t+1th iteration, is the particle velocity; The particle velocity update formula is: in, is the particle velocity of the t+1th iteration, ω is the inertia weight, c1, c2 are learning factors, r1, r2 are random numbers, p best is the historical optimal position of particle i, g best is the global optimal position; Calculate the fitness function J for each particle to determine its optimization effect. The determination formula is: Among them, J i represents the fitness value of particle i, w1η i is the combustion efficiency contribution value of particle i, and the weight w1 controls the priority of efficiency. represents the emission index penalty value of the particle, represents the equipment life risk penalty value of the particle; When the number of iterations reaches the upper limit or the improvement of the global optimal solution is lower than the threshold, the optimal adjustment strategy is output.

8. A boiler energy efficiency intelligent management method according to claim 7, characterized in that: The steps of transmitting the adjustment strategy to the burner, the denitration device and the water treatment equipment, and adjusting the air distribution strategy, the ammonia injection amount and the operating parameters of the circulation pump of the burner in real time are as follows: Decode the adjustment strategy and set the optimal parameter of the adjustment strategy output as P opt : P opt Decoded into device control instructions, including: Control the fuel injection rate and air distribution of the burner, Adjust the amount of ammonia sprayed by the denitrification device. Control the amount of water treatment agent added to the circulation pump of water treatment equipment; Calculate actual equipment control parameters, The fuel and air ratio of the burner is: Where φ is the fuel-to-air equivalence ratio, The ammonia spray concentration is adjusted to: in, For denitrification, spray ammonia concentration, is the flue gas velocity, measured by the flow meter, The concentration of water treatment agent is: Among them, C water is the concentration of water treatment agent, is the circulating water flow rate; The step of adjusting the air distribution strategy, ammonia injection amount and circulating pump operating parameters of the burner in real time also includes: the real-time controller should adjust the strategy, Controlling fuel injection rate using a burner controller and air flow Perform air distribution optimization, Use denitrification controller to adjust ammonia injection rate Maintain the activity range of the denitrification catalyst, Use water treatment controller to control the amount of water treatment agent added to the circulation pump Maintain stable water quality; Each controller receives sensor feedback data and performs real-time correction according to the adjusted operating parameters. The feedback mechanism formula is: k =u k-1 +K e ·(y set -y k ), where u k is the control parameter after the kth adjustment, y set is the target value, y k is the current feedback value, K e is the feedback gain.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the boiler energy efficiency intelligent management method described in any one of claims 1 to 8 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of a boiler energy efficiency intelligent management method according to any one of claims 1 to 8 are implemented.

Images