Intelligent biomass blending combustion method and device for coal-fired power plant

Abstract

The invention relates to the technical field of thermal power generation, and provides an intelligent biomass blending combustion method and device for a coal-fired power plant. The method comprises the steps of obtaining multi-source data for a coal-fired power plant, wherein the multi-source data comprises equipment inherent parameters, external instruction and demand data, fuel attribute parameters and real-time operation data; according to the multi-source data and a preset operation boundary constraint, calculating a target biomass blending combustion ratio and a corresponding target operation parameter through a multi-target dynamic optimization model; according to the target biomass blending combustion proportion and the corresponding target operation parameters, the combustion process of the coal-fired power plant is controlled through a feedforward-feedback composite control strategy. According to the method, the global optimization of the biomass blending combustion process of the coal-fired power plant is realized by constructing a data-driven intelligent decision-making and control closed loop. And the power generation cost and pollutant emission can be remarkably reduced, and the maximization of comprehensive benefits of safety, economy and environmental protection of a coal-fired power plant is realized.

Description

Technical Field

[0001] This invention relates to the field of thermal power generation technology, and in particular to a method and apparatus for intelligent co-firing of biomass in coal-fired power plants. Background Technology

[0002] To address the pressure of carbon emission reduction, co-firing biomass power generation with coal-fired power plants has become an important technological approach. Currently, the mainstream technology is direct co-combustion of biomass and coal, but significant problems exist in actual operation: First, biomass sources are diverse (such as straw, sawdust, rice husks, etc.), and its physicochemical properties, such as moisture, ash content, calorific value, and particle size, fluctuate dramatically, posing a significant challenge to stable boiler combustion, thermal efficiency, and pollutant emission control. Furthermore, the current co-combustion ratio is mostly set by operators based on experience, lacking scientific data support and optimization models. Second, the co-combustion process lacks precise metering and closed-loop control, resulting in poor uniformity and affecting combustion stability. Third, the operational objectives are singular, failing to synergistically optimize multiple goals such as fuel cost, carbon revenue, and environmental indicators. Therefore, there is an urgent need for an intelligent co-combustion method that can achieve optimal safety, economy, and environmental protection. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to provide a method and apparatus for intelligent co-firing of biomass in coal-fired power plants, so as to solve the above-mentioned technical problem.

[0004] The technical solution of this invention to solve the above-mentioned technical problems is as follows: A method for intelligent biomass blending in coal-fired power plants, comprising: acquiring multi-source data for the coal-fired power plant, wherein the multi-source data includes inherent equipment parameters, external command and demand data, fuel attribute parameters, and real-time operating data; calculating a target biomass blending ratio and corresponding target operating parameters through a multi-objective dynamic optimization model based on the multi-source data and preset operating boundary constraints; and controlling the combustion process of the coal-fired power plant through a feedforward-feedback composite control strategy based on the target biomass blending ratio and corresponding target operating parameters.

[0005] The beneficial effects of this invention are as follows: By constructing a data-driven intelligent decision-making and control closed loop, this invention achieves global optimization of the biomass co-firing process in coal-fired power plants. This method comprehensively utilizes multi-source data, solves for the optimal co-firing scheme through a multi-objective dynamic optimization model, and then ensures the precise execution of optimization commands through a feedforward-feedback composite control strategy. This significantly reduces power generation costs and pollutant emissions, maximizing the comprehensive safety, economic, and environmental benefits of coal-fired power plants.

[0006] Based on the above technical solution, the present invention can be further improved as follows.

[0007] Furthermore, the fuel attribute parameters include biomass characteristic data, which is obtained through the following methods: acquiring the current biomass procurement batch information; retrieving characteristic data matching the procurement batch information from a pre-built fuel characteristic database as estimated characteristic data, wherein the pre-built fuel characteristic database includes characteristic data and procurement information of biomass delivered to the plant in previous years; acquiring coal mill operating parameters; and dynamically correcting the estimated characteristic data based on the coal mill operating parameters to obtain the biomass characteristic data.

[0008] Furthermore, the multi-objective dynamic optimization model includes an economic and environmental benefit optimization layer and a combustion process optimization layer. The step of calculating the target biomass blending ratio and corresponding target operating parameters through the multi-source data and preset operating boundary constraints via the multi-objective dynamic optimization model includes: calculating the target biomass blending ratio using the economic and environmental benefit optimization layer as the objective function, based on the multi-source data and preset operating boundary constraints; and establishing a nonlinear mapping relationship between operating parameters and combustion results through the combustion process optimization layer, using the target biomass blending ratio and maximizing boiler efficiency as the objective function, and performing optimization to generate the target operating parameters.

[0009] Furthermore, the step of calculating the target biomass blending ratio based on the multi-source data and preset operational boundary constraints, with the objective function of maximizing comprehensive benefits, through the economic and environmental benefit optimization layer, includes: obtaining a feasible range for the blending ratio; generating multiple candidate blending ratios within the feasible range; for each candidate blending ratio, calculating a comprehensive economic benefit score corresponding to the candidate blending ratio based on the multi-source data and the predicted values ​​of boiler efficiency, pollutant emission concentration, and furnace temperature corresponding to the candidate blending ratio calculated by the combustion process optimization layer; and selecting the candidate blending ratio with the highest comprehensive economic benefit score from all candidate blending ratios that satisfy the operational boundary constraints as the target biomass blending ratio.

[0010] Furthermore, the step of generating the target operating parameters by establishing a nonlinear mapping relationship between operating parameters and combustion results and optimizing the process through the combustion process optimization layer, based on the target biomass co-firing ratio and with the goal of maximizing boiler efficiency as the objective function, includes: obtaining the adjustable range of operating parameters; and within the adjustable range of operating parameters, searching for the optimal combination of operating parameters based on a pre-trained boiler efficiency prediction model and a derivative-free optimization algorithm to obtain the target operating parameters.

[0011] Furthermore, the biomass characteristic data includes fuel moisture data and fuel heavy metal content; the control of the combustion process of the coal-fired power plant through a feedforward-feedback composite control strategy based on the target biomass blending ratio and corresponding target operating parameters includes: calculating primary air parameters based on the fuel moisture data, and adjusting the opening of the hot and cold air dampers of the primary air system according to the primary air parameters to stabilize the coal mill outlet temperature; calculating a corrosion risk index based on the fuel heavy metal content; determining an upper limit for the blending ratio based on the corrosion risk index; and generating an air distribution adjustment strategy based on the fuel heavy metal content, so as to control the combustion process of the coal-fired power plant according to the air distribution adjustment strategy and the upper limit for the blending ratio.

[0012] Furthermore, the control of the combustion process of the coal-fired power plant through a feedforward-feedback composite control strategy based on the target biomass blending ratio and corresponding target operating parameters further includes: using the speed of the biomass feeder and the speed of the coal feeder as control variables, and the actual blending ratio obtained in real time as the controlled variable, employing a PID control algorithm to coordinate and control the speed of the biomass feeder and the speed of the coal feeder so that the actual blending ratio approaches the target biomass blending ratio; using the opening degree of the secondary dampers at each level as control variables, and the nitrogen oxide concentration and fly ash carbon content measured in real time as the controlled variables, employing a multivariate model predictive control algorithm to dynamically adjust the opening degree of the secondary dampers at each level so that the nitrogen oxide concentration approaches the concentration set value and the fly ash carbon content approaches the carbon content set value.

[0013] Furthermore, the method also includes: obtaining the actual value of a preset operating indicator based on the multi-source data; calculating the deviation between the actual value of the preset operating indicator and the model prediction value to obtain a deviation value, wherein the model prediction value is the predicted value of the preset operating indicator calculated by the pre-trained boiler efficiency prediction model; comparing the deviation value with a deviation threshold; and performing parameter self-correction on the pre-trained boiler efficiency prediction model when the deviation value is continuously greater than the deviation threshold.

[0014] Furthermore, the method also includes: acquiring actual fuel sample analysis data and operational performance data for the current blending cycle; wherein, the actual fuel sample analysis data is data describing the physicochemical properties of the fuel actually used, and the operational performance data is data describing the combustion response of the boiler system to the fuel actually used; associating the actual fuel sample analysis data and the operational performance data, and recording the associating data into the pre-constructed fuel characteristic database.

[0015] To address the aforementioned technical problems, the present invention also provides an intelligent biomass co-firing device for coal-fired power plants, comprising: The data acquisition module is used to acquire multi-source data for coal-fired power plants, including inherent equipment parameters, external command and demand data, fuel attribute parameters, and real-time operation data. The decision module is used to calculate the target biomass co-firing ratio and the corresponding target operating parameters based on the multi-source data and preset operating boundary constraints through a multi-objective dynamic optimization model. The execution module is used to control the combustion process of the coal-fired power plant according to the target biomass blending ratio and the corresponding target operating parameters through a feedforward-feedback composite control strategy. Attached Figure Description

[0016] Figure 1 This is a flowchart of a method for intelligent biomass co-firing in a coal-fired power plant according to the present invention; Figure 2 This is a flowchart of the method according to Embodiment 1 of the present invention; Figure 3 This is a flowchart of the multi-objective optimization strategy of the present invention; Figure 4 This is a schematic diagram of an intelligent biomass co-firing device for a coal-fired power plant according to the present invention. Detailed Implementation

[0017] The principles and features of the present invention are described below. The examples given are only for explaining the present invention and are not intended to limit the scope of the present invention.

[0018] Example 1 like Figure 1 As shown, this embodiment provides a method for intelligent co-firing of biomass in a coal-fired power plant, including: S101. Obtain multi-source data for coal-fired power plants, including equipment inherent parameters, external command and demand data, fuel attribute parameters, and real-time operation data.

[0019] S102. Based on multi-source data and preset operational boundary constraints, calculate the target biomass co-firing ratio and corresponding target operating parameters through a multi-objective dynamic optimization model.

[0020] S103. Based on the target biomass blending ratio and the corresponding target operating parameters, the combustion process of the coal-fired power plant is controlled through a feedforward-feedback composite control strategy.

[0021] This invention achieves global optimization of the biomass co-firing process in coal-fired power plants by constructing a data-driven intelligent decision-making and control closed loop. The method comprehensively utilizes multi-source data, solves for the optimal co-firing scheme through a multi-objective dynamic optimization model, and then ensures the precise execution of optimization commands through a feedforward-feedback composite control strategy. This significantly reduces power generation costs and pollutant emissions, maximizing the comprehensive safety, economic, and environmental benefits of coal-fired power plants.

[0022] Optionally, in an embodiment, the fuel attribute parameters include biomass characteristic data, which is obtained by: acquiring the current biomass procurement batch information; retrieving characteristic data matching the procurement batch information from a pre-built fuel characteristic database as estimated characteristic data, the pre-built fuel characteristic database including characteristic data and procurement information of biomass delivered to the plant in previous years; acquiring coal mill operating parameters; and dynamically correcting the estimated characteristic data based on the coal mill operating parameters to obtain the biomass characteristic data.

[0023] like Figure 2 As shown, the construction of the fuel characteristic database specifically involves: continuously collecting historical biomass fuel characteristic data received from the plant, including received moisture (Mar), dry ash-free volatile matter (Vdaf), received lower heating value (Qnet, ar), ash content (Ad), carbon content (C), chlorine content (Cl), alkali metal content (K, Na), etc., to form a continuously updated comprehensive database, which is then linked with information such as procurement source, season, and price to form biomass fuel data assets.

[0024] Based on the procurement batch information (season, price, source, etc.) of the currently operational biomass, characteristic data for that source or category is retrieved from the fuel characteristic database as a prediction of the current fuel characteristics. The characteristic data mainly requires the composition information of the biomass fuel, which is strongly correlated with the place of origin. For biomass of the same category from the same region, the characteristic data will not differ much.

[0025] Based on the operating parameters of the coal mill (such as outlet temperature, current, primary air volume, etc.), key changing parameters such as the overall moisture content of the biomass currently entering the furnace are deduced to achieve dynamic approximation of fuel characteristics. The deduction process mainly relies on energy balance and mass balance, and the experience of operators allows for a general judgment. The specific principle is as follows: During operation, parameters such as primary air volume, outlet temperature, and current of the coal mill are closely related to the moisture content of the fuel. The higher the moisture content of the fuel, the more heat is required for evaporation, resulting in a decrease in outlet temperature. The primary air volume affects the drying capacity, and changes in air volume can reflect the moisture evaporation. The coal mill current is related to the grinding load, and changes in fuel moisture may affect the current value (for example, increased moisture usually leads to an increase in current because the fuel is more viscous and difficult to grind).

[0026] First, information such as the outlet temperature, primary air volume, primary air temperature, current, total fuel mass flow rate, biomass feed rate, coal feed rate, and coal moisture content of the coal mill are collected. The moisture content of the mixed fuel is calculated through the heat balance of the coal mill. Then, the moisture content of the biomass is calculated based on the known moisture content of the coal and dynamically adjusted accordingly.

[0027] Specifically, the inherent parameters of the equipment include: Boiler design parameters include: furnace structure, heating surface arrangement, and burner type; equipment performance curves: performance curves of major equipment such as coal mill and fans.

[0028] External instructions and demand data, including: Grid side: power grid dispatch instructions, heating demand (if any).

[0029] Fuel property parameters include: source-based: coal / biomass fuel characteristics (including as-received moisture (Mar), dry ash-free volatile matter (Vdaf), as-received lower heating value (Qnet, ar), ash content (Ad), carbon content (C), chlorine content (Cl), alkali metal content (K, Na), etc.), that is, biomass characteristic data and coal characteristic data.

[0030] Real-time operating data includes: furnace temperature, flue gas parameters (O2, CO, NO) x (SO2), main steam pressure / temperature, carbon content of fly ash / slag, air-coal ratio, etc.

[0031] Optionally, in this embodiment, the multi-objective dynamic optimization model includes an economic and environmental benefit optimization layer and a combustion process optimization layer. Based on multi-source data and preset operating boundary constraints, the multi-objective dynamic optimization model calculates the target biomass blending ratio and the corresponding target operating parameters, including: based on multi-source data and preset operating boundary constraints, using the maximization of comprehensive benefits as the objective function, calculating the target biomass blending ratio through the economic and environmental benefit optimization layer; and based on the target biomass blending ratio, using the maximization of boiler efficiency as the objective function, establishing a nonlinear mapping relationship between operating parameters and combustion results through the combustion process optimization layer and performing optimization to generate the target operating parameters.

[0032] like Figure 3 As shown, a multi-objective dynamic optimization model is formulated, employing hierarchical optimization. The upper layer is the economic and environmental benefit optimization layer, with maximizing comprehensive benefits as the primary objective. It comprehensively considers power generation revenue, carbon emission reduction revenue, fuel costs, and environmental costs, maximizing [power generation revenue + carbon emission reduction revenue - (coal cost + biomass cost + environmental penalty cost)]. It also considers operational safety, equipment, and environmental constraints, outputting the biomass co-firing ratio.

[0033] The lower layer is the combustion process optimization layer. With the safe and stable operation of the boiler as the fundamental constraint, the objective function is to maximize the boiler efficiency. The algorithm is to use deep learning network algorithm or reinforcement learning algorithm to establish a nonlinear mapping relationship between the operating parameters and the combustion results, and to optimize and output the target operating parameters, including the air distribution method, coal mill outlet air temperature / volume, etc.

[0034] Optionally, in this embodiment, based on multi-source data and preset operational boundary constraints, and with the goal of maximizing comprehensive benefits, the target biomass blending ratio is calculated through an economic and environmental benefit optimization layer. This includes: obtaining a feasible range for the blending ratio; generating multiple candidate blending ratios within the feasible range; for each candidate blending ratio, calculating a comprehensive economic benefit score based on multi-source data and the predicted values ​​of boiler efficiency, pollutant emission concentration, and furnace temperature corresponding to the candidate blending ratio calculated by the combustion process optimization layer; and selecting the candidate blending ratio with the highest comprehensive economic benefit score from all candidate blending ratios that meet the operational boundary constraints as the target biomass blending ratio.

[0035] Specifically, the revenue from power generation can be calculated using the following formula: P_power×(LHV_coal×M_coal+LHV_bio×M_bio)×η_boiler(r); Where P_power is the grid-connected electricity price (yuan / kWh); LHV is the lower heating value of the fuel (kWh / kg); and M is the fuel mass flow rate (kg / h).

[0036] The formula for calculating M_bio is: M_bio = r × M_total.

[0037] The formula for calculating M_coal is: M_coal=(1-r)×M_total.

[0038] η_boiler(r) is the key coupling point between the upper and lower layers. The upper-layer model itself cannot accurately calculate η_boiler; it needs feedback from the lower-layer model of an estimated or measured boiler efficiency value corresponding to the current r (co-firing ratio).

[0039] The benefits of carbon emission reduction can be calculated using the following formula: (M_bio×EF_coal-M_bio×EF_bio)×P_carbon; Where EF is the carbon emission factor (kgCO2 / kg fuel); P_carbon is the carbon trading price (yuan / kgCO2). It is generally believed that CO2 emissions from biomass combustion are zero, so this item can be simplified to M_bio×EF_coal×P_carbon.

[0040] Fuel costs can be calculated using the following formula: P_coal×M_coal+P_bio×M_bio; Wherein, P_coal and P_bio are the unit price of fuel (yuan / kg).

[0041] The cost of environmental penalties can be calculated using the following formula: α×max(0,NOx(r)-NOx_lim)+β×max(0,SOx(r)-SOx_lim)+...; Where α, β: penalty coefficients (yuan / mg / m³) 3 ); NOx(r), SOx(r): The upper-level model needs to obtain the estimated pollutant emission concentrations given r from the lower-level model.

[0042] The constraints specifically include: Safety constraints apply, for example, T_furnace(r) ≤ T_max. T_furnace(r) (furnace temperature) also requires predictions from lower-level models. In the optimization algorithm, this is treated as a "hard constraint," and any violation of this constraint will result in the immediate rejection of any possible value for r.

[0043] Equipment capacity constraints: For example, r_min ≤ r ≤ r_max. r_max depends not only on the inventory, but also on the maximum biomass processing capacity of the coal mill and the feeding system.

[0044] Environmental constraints: For example, NOx(r) ≤ NOx_lim. This is the basis for the penalty term in the objective function and may also be used as a hard constraint.

[0045] The specific solution logic (constrained single-variable nonlinear programming, where the variable is the blending ratio) is as follows: Inputs: real-time electricity price, carbon price, fuel characteristics, price, etc.; the latest mapping or model of the relationship between boiler efficiency, pollutant emissions, etc., obtained from the lower-level algorithm.

[0046] Solution: A precise linear search method can be used. Within the feasible interval [r_min, r_max] of r, a series of candidate points r_candidate are generated with a certain step size (e.g., 0.5%). For each r_candidate, the relational model passed from the lower layer is called to estimate the corresponding key parameters such as efficiency, emissions, and furnace temperature. These parameters are substituted into the objective function to calculate the economic benefit score for each r_candidate. Simultaneously, each candidate point is rigorously checked to ensure it meets all hard constraints (e.g., furnace temperature does not exceed limits). All points that do not meet the constraints are eliminated. Among all candidate points that meet the constraints, the one with the largest objective function value is selected; its corresponding r is the optimal blending ratio r* for this optimization cycle.

[0047] Output: Pass r* as the set value to the lower-level algorithm.

[0048] Optionally, in an embodiment, based on the target biomass co-firing ratio and with maximizing boiler efficiency as the objective function, a nonlinear mapping relationship between operating parameters and combustion results is established and optimized through a combustion process optimization layer to generate target operating parameters, including: obtaining the adjustable range of operating parameters; within the adjustable range of operating parameters, searching for the optimal combination of operating parameters based on a pre-trained boiler efficiency prediction model and a derivative-free optimization algorithm to obtain the target operating parameters.

[0049] Specifically, the objective function of the lower layer (combustion process optimization) is: maximizing boiler efficiency.

[0050] Boiler efficiency η_boiler is a comprehensive indicator determined by combustion efficiency, heat transfer efficiency, mechanical incomplete combustion losses, and flue gas losses. It cannot be directly measured, but it can be calculated from measurable parameters such as oxygen content, flue gas temperature, carbon monoxide concentration, and fly ash carbon content. Lower-level optimization involves adjusting operating parameters to indirectly maximize this calculated efficiency value.

[0051] First, a deep learning network needs to be trained. This network is not used for direct optimization, but rather to build a dynamic proxy model of the combustion process. The logic is as follows: Input: X=[blending ratio r, load requirement, coal quality, biomass moisture, primary air volume, pulverizer outlet temperature, secondary air ratio, burnout air opening, O2 setpoint, ...]; Output: Y = [Predicted boiler efficiency, predicted NOx concentration, predicted furnace temperature, predicted CO concentration, ...]; Training data: Historical operating data from DCS, covering various operating conditions.

[0052] Logical function: Once the network is trained, given a set of operating parameters, it can instantly predict the combustion results (efficiency, emissions, etc.) and quickly evaluate the merits of a large number of operating schemes.

[0053] The detailed logic of the optimization process is as follows: Suppose that the upper level has issued an instruction: the blending ratio is r*.

[0054] Perceive the current status: Read the current uncontrollable factors such as load, fuel characteristics, and ambient temperature.

[0055] Define the operating space: Determine the reasonable range of adjustable parameters, such as: primary air volume: [PA_min, PA_max], coal mill outlet temperature: [T_out_min, T_out_max], secondary air ratio: [SA_ratio_min, SA_ratio_max], total air volume / O2 setpoint: [O2_setpoint_min, O2_setpoint_max].

[0056] Searching within the operation space: Since a deep learning model is a "black box" function that cannot be differentiated, derivativeless optimization algorithms are required, such as genetic algorithms, particle swarm optimization, or Bayesian optimization.

[0057] Taking particle swarm optimization as an example: a. Initialization: Randomly generate a group of "particles", the position of each particle represents a set of operating parameters [primary air volume, outlet temperature, secondary air ratio, O2 setting].

[0058] b. Evaluation: The position (operational parameters) of each particle is combined with fixed r*, current load, fuel characteristics, etc. into a complete input, which is then fed into a trained deep learning network. The network outputs the predicted boiler efficiency.

[0059] c. Update: Each particle updates its velocity and direction (i.e., adjusts its operating parameters) based on its own historical best position and the best position of the entire swarm.

[0060] d. Loop: Repeat steps b and c until the maximum number of iterations is reached or the result converges.

[0061] Output and execution: The optimal particle position obtained from the search, i.e. the optimal combination of operating parameters, is output to the underlying controller of the DCS (such as a PLC), which then executes the command and adjusts equipment such as dampers and coal feeders.

[0062] The feedforward-feedback composite control strategy includes feedforward control and feedback control. In this embodiment, two feedforward control methods are provided: adjusting the pulverizer temperature and airflow in advance based on fuel moisture changes to ensure drying output; and adjusting the blending ratio and air distribution method in advance based on fuel heavy metal content to mitigate corrosion risks. Two feedback control methods are also provided: using the feeder speed as the control variable and the real-time calculated blending ratio as the controlled variable to overcome feed fluctuations; and using the secondary damper opening as the control variable and the real-time NOx concentration and fly ash carbon content as the controlled variables for dynamic optimization.

[0063] Optionally, in this embodiment, the biomass characteristic data includes fuel moisture data and fuel heavy metal content; based on the target biomass blending ratio and the corresponding target operating parameters, the combustion process of the coal-fired power plant is controlled through a feedforward-feedback composite control strategy, including: calculating primary air parameters based on fuel moisture data, and adjusting the opening of the hot and cold air dampers of the primary air system according to the primary air parameters to stabilize the coal mill outlet temperature; calculating a corrosion risk index based on fuel heavy metal content; determining an upper limit for the blending ratio based on the corrosion risk index; and generating an air distribution adjustment strategy based on fuel heavy metal content, so as to control the combustion process of the coal-fired power plant according to the air distribution adjustment strategy and the upper limit for the blending ratio.

[0064] The feedforward control based on fuel moisture content specifically involves: Objective: To ensure that the outlet temperature of the coal mill is within the set range, to ensure that the fuel is fully dried, and to prevent excessive temperature from causing deflagration.

[0065] Controlled quantities: primary air hot damper opening (adjusting primary air temperature), cold damper opening (adjusting primary air temperature), primary air volume (controlled by primary air fan guide vanes or speed).

[0066] Controlled variable: Coal mill outlet temperature.

[0067] Control Principle: Increased fuel moisture content requires more heat to evaporate the moisture; otherwise, the outlet temperature will drop, potentially leading to mill blockage and unstable combustion. By detecting (or estimating) the overall fuel moisture content in advance, the required primary air temperature or volume is calculated, and feedforward compensation is implemented.

[0068] The specific calculation steps are as follows: Establish a thermal balance model for the coal mill: The heat carried in by the primary air = heat required to evaporate moisture + heat required to heat fuel to the outlet temperature + heat loss, i.e., Q_air = Q_evaporation + Q_heating_fuel + Q_loss; Where, Q_air = m_air × c_p_air × (T_in - T_ref) [heat carried in by the primary wind]; Q_evaporation = m_water × (h_vap + c_p_vapor × (T_out - T_ref)) [Heat required to evaporate water]; Q_heating_fuel = m_fuel_dry × c_p_fuel × (T_out - T_fuel_in) [Heat required to heat dry fuel]; m_water = m_fuel_total × w[water mass flow rate, w is wet basis water].

[0069] Based on the target outlet temperature T_out_set, the required primary air temperature T_in_required is calculated in reverse: The heat balance equation is: m_air×c_p_air×(T_in_required-T_ref)=m_fuel_total×w×(h_vap+c_p_vapor×(T_out_set-T_ref))+m_fuel_total×(1-w)×c_p_fuel×(T_out_set-T_fuel_in)+Q_loss.

[0070] Then, adjust the opening of the hot air damper and cold air damper according to T_in_required to ensure that the primary air temperature reaches T_in_required. At the same time, ensure that the primary air volume also meets the requirements for conveying and drying. The primary air volume usually has a lower limit to prevent pipe blockage.

[0071] In addition, if the primary air temperature has reached its maximum, but the calculated T_in_required is higher, it may be necessary to reduce the feed rate (reduce the load) or increase the primary air volume (but note that increasing the air volume will reduce the outlet temperature, so a comprehensive balance must be made).

[0072] In actual control, the feedforward control calculates the primary air temperature setpoint based on moisture changes, and then the primary air temperature control system (by adjusting the cold and hot air dampers) tracks this setpoint. Simultaneously, the coal mill outlet temperature controller acts as feedback correction to eliminate feedforward errors.

[0073] The feedforward control based on heavy metal content specifically involves: Objective: To mitigate the corrosion risk of boiler heating surfaces. Heavy metals (such as potassium, sodium, and chlorine) form low-melting-point compounds during combustion, which adhere to the heating surfaces and cause high-temperature corrosion.

[0074] Controlled quantities: blending ratio, air distribution method (such as secondary air ratio, burnout air location, etc.).

[0075] Controlled variable: There is no directly controlled variable; it belongs to feedforward compensation.

[0076] Control principle: When a high level of heavy metals is detected in the fuel, the total input of heavy metals is reduced by decreasing the blending ratio (reducing biomass input). At the same time, corrosion is mitigated by adjusting the air distribution method, such as increasing the air volume in the combustion zone (lowering the combustion temperature and reducing heavy metal gasification) or changing the atmosphere in the combustion zone (a reducing atmosphere will aggravate corrosion, so ensuring sufficient oxygen is crucial).

[0077] The specific calculation steps are as follows: Establish a mapping model between heavy metal content and corrosion risk. This model can be an empirical model, for example: Corrosion risk index = k1×[K] + k2×[Na] + k3×[Cl] + ... [where [K] represents the concentration].

[0078] The upper limit of the blending ratio is determined based on the corrosion risk index. For example: If the corrosion risk index is greater than the threshold, then the co-firing ratio r ≤ r_max_safe (safe upper limit, determined by experiment or experience).

[0079] Simultaneously, based on the heavy metal content and blending ratio, an adjustment strategy for the air distribution method is determined. For example, increasing the secondary air volume to create an oxidizing atmosphere in the combustion zone and avoid a reducing atmosphere; adjusting the combustion zone temperature to avoid localized high temperatures and reduce the volatilization of heavy metals.

[0080] Feedforward control based on heavy metal content does not rely on online corrosion measurement (which is difficult to measure in real time), but rather adjusts in advance based on fuel analysis data.

[0081] Optionally, in the embodiments, the combustion process of the coal-fired power plant is controlled by a feedforward-feedback composite control strategy based on the target biomass blending ratio and the corresponding target operating parameters. This further includes: using the speed of the biomass feeder and the speed of the coal feeder as control variables, and the actual blending ratio obtained in real time as the controlled variable, employing a PID control algorithm to coordinate and control the speed of the biomass feeder and the speed of the coal feeder so that the actual blending ratio approaches the target biomass blending ratio; and using the opening degree of each layer of secondary dampers as control variables, and the nitrogen oxide concentration and fly ash carbon content measured in real time as the controlled variables, employing a multivariate model predictive control algorithm to dynamically adjust the opening degree of each layer of secondary dampers so that the nitrogen oxide concentration approaches the set concentration value and the fly ash carbon content approaches the set carbon content value.

[0082] The specific method for controlling the blending ratio is as follows: Objective: To overcome feed fluctuations and stabilize the actual blending ratio at the set value (given by the upper-level algorithm).

[0083] Controlled parameters: rotational speed of biomass feeder and rotational speed of coal feeder.

[0084] Controlled quantity: The blending ratio r = M_bio / (M_bio+M_coal) calculated in real time.

[0085] Control principle: By adjusting the speed of the two feeders, the feed rates of biomass and coal are controlled, thereby maintaining the blending ratio.

[0086] The specific calculation steps are as follows: Real-time calculation of blending ratio: r_actual=M_bio_actual / (M_bio_actual+M_coal_actual).

[0087] The actual blending ratio is compared with the set value r_set to obtain the error e = r_set - r_actual.

[0088] A PID controller is used to calculate the control quantity (adjustment of feeder speed) based on the error e: u_bio = PID(e) [Adjustment amount for biomass feeder speed]; u_coal=-PID(e)×(M_coal_nominal / M_bio_nominal)[Adjustment amount of coal feeder speed, note that the two are in opposite directions].

[0089] The two feeders need to be coordinated to avoid excessive fluctuations in the total fuel quantity. Alternatively, a master-slave control strategy can be adopted, for example, with the total fuel quantity as the master control and the blending ratio as the slave control.

[0090] The feed rate can be estimated by the feeder speed, but it is best to have real-time feed rate measurement (such as a weighing feeder). Otherwise, it is necessary to calibrate the relationship between the feeder speed and the feed rate.

[0091] The specific methods of combustion optimization control are as follows: Objective: To reduce NOx emissions while ensuring combustion efficiency.

[0092] Controlled quantity: Secondary air damper opening (including main combustion zone damper, burnout damper, etc.).

[0093] Controlled quantities: NOx concentration (measured by a flue gas analyzer) and fly ash carbon content (obtainable by online monitoring equipment or soft measurement).

[0094] Control principle: There is a coupling relationship between NOx and fly ash carbon content: measures to reduce NOx (such as staged air distribution) may lead to incomplete combustion and increased fly ash carbon content. Therefore, it is necessary to find a balance point that minimizes fly ash carbon content while keeping NOx levels within acceptable limits.

[0095] The control strategy is as follows: Multivariate optimization problem: NOx concentration and fly ash carbon content can be treated as two controlled variables, and secondary damper opening (multiple) as control variables, to construct a multiple-input multiple-output (MIMO) control system. Due to the complexity of the combustion process and its large delay and nonlinearity, advanced control algorithms such as model predictive control (MPC) can be used.

[0096] Taking model predictive control as an example: A dynamic model (which can be a transfer function model or a state-space model) is established to correlate NOx and fly ash carbon content with the secondary damper opening. In each control cycle, the controller predicts the NOx and fly ash carbon content over a future period based on current measurements and future setpoints. An optimization algorithm is used to calculate the adjustment sequence of the secondary damper opening over this future period, ensuring that the predicted output is as close as possible to the setpoint while satisfying operational constraints. The first calculated control variable is then applied to the system.

[0097] The determination of the set values ​​is as follows: the set value of NOx is usually determined by environmental standards, while the set value of carbon content in fly ash is determined by economic factors (the lower the better, but subject to NOx constraints).

[0098] Optionally, in an embodiment, the method further includes: obtaining the actual value of a preset operating indicator based on multi-source data; calculating the deviation between the actual value of the preset operating indicator and the model prediction value to obtain a deviation value, wherein the model prediction value is the predicted value of the preset operating indicator calculated by the pre-trained boiler efficiency prediction model; comparing the deviation value with a deviation threshold; and performing parameter self-correction on the pre-trained boiler efficiency prediction model when the deviation value is continuously greater than the deviation threshold.

[0099] Continuously monitor the actual combustion results (efficiency, emissions), compare the actual data with the predicted data of the deep learning model. If the deviation is consistently large, it indicates that the change in operating conditions has led to a decrease in model accuracy. It is necessary to activate the online model update mechanism and fine-tune the deep learning network with new data to keep it synchronized with the actual boiler.

[0100] Actual operational performance can be judged by selecting operational indicators, specifically boiler efficiency, NOx emissions, coal mill outlet moisture content, and furnace temperature. Furthermore, deviations are assessed using a combination of relative and absolute deviation methods.

[0101] For efficiency-proportional parameters, use relative deviation: Deviation_n=(n_predicted-n_actual) / n_actual×100%.

[0102] For concentration-related parameters such as NOx, use absolute deviation: Deviation_NOx=NOx_predicted-NOx_actual.

[0103] Comprehensive Deviation Index (Weighted Summary): Overall_Deviation=w1×|Deviation_η|+w2×|Deviation_NOx|+w3×|Deviation_T_out|.

[0104] The deviation threshold can vary depending on the operating conditions. Under high load conditions, the efficiency deviation threshold is 0.5, and the concentration deviation threshold is 5.0 (mg / m³). 3 Under low load conditions, the deviation threshold is amplified, with an efficiency deviation threshold of 1.0 and a concentration deviation threshold of 8.0.

[0105] The persistence of a problem is mainly judged from a time-related perspective. For example, for coal mill outlet temperature and air volume regulation, which have short response times, adjustments are needed if the deviation lasts for more than 2 minutes. For NOx concentration and furnace temperature, which have minute-level responses, the judgment time becomes 15 or 20 minutes or more. For boiler efficiency, which has an even longer response time, the judgment is based on hours or more.

[0106] Optionally, in an embodiment, the method further includes: acquiring actual fuel sample analysis data and operational performance data for the current blending cycle; wherein, the actual fuel sample analysis data is data describing the physicochemical properties of the fuel actually used, and the operational performance data is data describing the combustion response of the boiler system to the fuel actually used; the actual fuel sample analysis data and operational performance data are correlated, and the correlated data is entered into a pre-built fuel characteristic database.

[0107] The actual operating results are compared with the predicted values ​​of the optimized model. If the deviation persists, the model parameter self-correction is initiated. Simultaneously, the final actual fuel sample analysis data from this blending cycle is correlated with the operating results and then entered into the fuel characteristic database to enrich its content and make subsequent fuel characteristic characterization more accurate. The actual fuel sample analysis data refers to the characteristic data of the actual fuel fed into the furnace, obtained through experimental analysis and other methods.

[0108] In summary, this method addresses the shortcomings of biomass blending processes in coal-fired power plants. By establishing a fuel characteristic database and combining it with a multi-objective optimization algorithm, it provides a data-driven intelligent biomass blending method that achieves adaptive adaptation to fluctuations in biomass fuel characteristics and global optimization of the blending process. This method mainly includes the following aspects: First, constructing a fuel characteristic database, collecting and inputting fuel characteristic data of biomass entering the plant over the years to form data assets; second, characterizing the characteristics of the fuel entering the furnace, analyzing the current biomass fuel characteristics based on the fuel characteristic database; third, multi-objective optimization decision-making, specifying reasonable evaluation indicators and selecting appropriate constraints to optimize the operation process; and fourth, precise control, combining intelligent algorithms to select appropriate feedback signals and dynamically adjust and optimize input values.

[0109] Based on this, the method has the following advantages: 1. This method precisely optimizes the co-firing ratio, maximizes the utilization of low-cost biomass and carbon emission reduction benefits, and reduces power generation costs.

[0110] 2. This method uses a historical fuel characteristic database as the basis for decision-making. Through a self-learning mechanism, it adapts to the fluctuations in biomass fuel characteristics and exhibits strong control robustness.

[0111] 3. This method adopts a collaborative self-learning approach between the database and the optimization model. The database provides training data for the model, and the effective data generated by the model optimizes the database, forming a virtuous cycle of mutual promotion between "data and model".

[0112] Example 2 like Figure 4 As shown, this embodiment provides a smart biomass co-firing device 200 for coal-fired power plants, comprising: Data acquisition module 201 is used to acquire multi-source data for coal-fired power plants. The multi-source data includes inherent equipment parameters, external command and demand data, fuel attribute parameters, and real-time operation data. The decision module 202 is used to calculate the target biomass co-firing ratio and the corresponding target operating parameters based on multi-source data and preset operating boundary constraints through a multi-objective dynamic optimization model. The execution module 203 is used to control the combustion process of the coal-fired power plant through a feedforward-feedback composite control strategy based on the target biomass co-firing ratio and the corresponding target operating parameters.

[0113] Optionally, in an embodiment, the fuel attribute parameters include biomass characteristic data, which is obtained by: acquiring the current biomass procurement batch information; retrieving characteristic data matching the procurement batch information from a pre-built fuel characteristic database as estimated characteristic data, the pre-built fuel characteristic database including characteristic data and procurement information of biomass delivered to the plant in previous years; acquiring coal mill operating parameters; and dynamically correcting the estimated characteristic data based on the coal mill operating parameters to obtain the biomass characteristic data.

[0114] Optionally, in this embodiment, the multi-objective dynamic optimization model includes an economic and environmental benefit optimization layer and a combustion process optimization layer; the decision module 202 includes: The upper-level computing unit is used to calculate the target biomass co-firing ratio based on multi-source data and preset operational boundary constraints, with the objective function of maximizing comprehensive benefits, through the economic and environmental benefits optimization layer. The lower-level calculation unit is used to generate target operating parameters by establishing a nonlinear mapping relationship between operating parameters and combustion results through the combustion process optimization layer, with the objective function of maximizing boiler efficiency, based on the target biomass co-firing ratio.

[0115] Optionally, in this embodiment, the upper-layer computing unit includes: The interval acquisition sub-unit is used to obtain the feasible interval of the blending ratio; The ratio generation subunit is used to generate multiple candidate blending ratios within the feasible range of blending ratios. The scoring calculation subunit is used to calculate the comprehensive economic benefit score corresponding to each candidate blending ratio based on multi-source data and the predicted values ​​of boiler efficiency, pollutant emission concentration and furnace temperature corresponding to the candidate blending ratio calculated by the combustion process optimization layer. The target ratio determination sub-unit is used to select the candidate co-firing ratio with the highest comprehensive economic benefit score from all candidate co-firing ratios that meet the operational boundary constraints as the target biomass co-firing ratio.

[0116] Optionally, in an embodiment, the lower-level computing unit includes: The range acquisition subunit is used to acquire the adjustable range of the operation parameters; The optimization sub-unit is used to search for the optimal combination of operating parameters based on a pre-trained boiler efficiency prediction model and a derivative-free optimization algorithm within the adjustable range of operating parameters, so as to obtain the target operating parameters.

[0117] Optionally, in this embodiment, the biomass characteristic data includes fuel moisture data and fuel heavy metal content; the execution module 203 includes: The first feedforward unit is used to calculate the primary air parameters based on the fuel moisture data, and adjust the opening of the hot air damper and cold air damper of the primary air system according to the primary air parameters to stabilize the coal mill outlet temperature. Risk calculation unit, used to calculate corrosion risk index based on fuel heavy metal content; The upper limit calculation unit is used to determine the upper limit of the blending ratio based on the corrosion risk index. The second feedforward unit is used to generate an air distribution mode adjustment strategy based on the heavy metal content of the fuel, so as to control the combustion process of the coal-fired power plant according to the air distribution mode adjustment strategy and the upper limit of the blending ratio.

[0118] Optionally, in an embodiment, the execution module 203 further includes: The first feedback unit is used to coordinate the control of the biomass feeder speed and the coal feeder speed with the actual co-firing ratio obtained in real time as the controlled variable, and adopts the PID control algorithm to coordinate the control of the biomass feeder speed and the coal feeder speed so that the actual co-firing ratio approaches the target biomass co-firing ratio. The second feedback unit is used to dynamically adjust the opening of the secondary dampers at each level, with the opening of the secondary dampers at each level as the control variable and the nitrogen oxide concentration and fly ash carbon content measured in real time as the controlled variables. It employs a multivariate model predictive control algorithm to adjust the opening of the secondary dampers at each level so that the nitrogen oxide concentration approaches the set concentration value and the fly ash carbon content approaches the set carbon content value.

[0119] Optionally, in an embodiment, the apparatus further includes: The actual value acquisition module is used to obtain the actual values ​​of preset operating indicators based on multi-source data; The deviation calculation module is used to calculate the deviation between the actual value and the model prediction value of the preset operating index, and obtain the deviation value. The model prediction value is the predicted value of the preset operating index calculated by the pre-trained boiler efficiency prediction model. The comparison module is used to compare the deviation value with the deviation threshold; The self-calibration module is used to perform parameter self-calibration on the pre-trained boiler efficiency prediction model when the deviation value continues to exceed the deviation threshold.

[0120] Optionally, in an embodiment, the apparatus further includes: The acquisition module is used to acquire actual fuel sample analysis data and operational performance data for the current blending cycle; Among them, the actual fuel sample analysis data describes the physicochemical properties of the fuel actually used, and the operating effect data describes the combustion response of the boiler system to the fuel actually used. The data association module is used to associate actual fuel sample analysis data and operational performance data, and then input the associated data into a pre-built fuel characteristic database.

[0121] In some embodiments, the intelligent biomass co-firing device 200 for coal-fired power plants of the present invention can be implemented in a combination of hardware and software. As an example, the intelligent biomass co-firing device 200 for coal-fired power plants of the present invention can be a processor in the form of a hardware decoding processor, which is programmed to execute the intelligent biomass co-firing method for coal-fired power plants of the present invention. For example, the processor in the form of a hardware decoding processor can be one or more application-specific integrated circuits (ASICs), DSPs, programmable logic devices (PLDs), complex programmable logic devices (CPLDs), field-programmable gate arrays (FPGAs), or other electronic components.

[0122] The modules described in the embodiments of this invention can be implemented in software or hardware. The names of the modules are not, in some cases, limiting the scope of the module itself.

[0123] The above description is merely a preferred embodiment of the present invention and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of disclosure in this invention is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-disclosed concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features with similar functions disclosed in this invention.

[0124] It should be noted that the terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and represent a limitation on a specific order or sequence. Where appropriate, the order of use for similar objects can be interchanged so that the embodiments of this application described herein can be implemented in an order other than that shown or described.

[0125] Those skilled in the art will recognize that this invention can be implemented as a system, method, or computer program product. Therefore, this invention can be specifically implemented in the following forms: it can be entirely hardware, entirely software (including firmware, resident software, microcode, etc.), or a combination of hardware and software, generally referred to herein as a "circuit," "module," or "system." Furthermore, in some embodiments, this invention can also be implemented as a computer program product contained in one or more computer-readable media, which includes computer-readable program code.

[0126] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for intelligent co-firing of biomass in a coal-fired power plant, characterized in that, include: Acquire multi-source data for coal-fired power plants, including inherent equipment parameters, external command and demand data, fuel attribute parameters, and real-time operating data; Based on the multi-source data and preset operational boundary constraints, the target biomass co-firing ratio and corresponding target operating parameters are calculated through a multi-objective dynamic optimization model. Based on the target biomass blending ratio and the corresponding target operating parameters, the combustion process of the coal-fired power plant is controlled through a feedforward-feedback composite control strategy.

2. The intelligent biomass co-firing method for coal-fired power plants according to claim 1, characterized in that, The fuel property parameters include biomass characteristic data, which is obtained through the following methods: Obtain information on the current biomass procurement batches; The characteristic data that matches the procurement batch information is retrieved from the pre-built fuel characteristic database as the estimated characteristic data. The pre-built fuel characteristic database includes characteristic data and procurement information of biomass delivered to the plant in previous years. Obtain the operating parameters of the coal mill; Based on the operating parameters of the coal mill, the estimated characteristic data is dynamically corrected to obtain the biomass characteristic data.

3. The intelligent biomass co-firing method for coal-fired power plants according to claim 1, characterized in that, The multi-objective dynamic optimization model includes an economic and environmental benefit optimization layer and a combustion process optimization layer; the calculation of the target biomass blending ratio and corresponding target operating parameters based on the multi-source data and preset operating boundary constraints through the multi-objective dynamic optimization model includes: Based on the multi-source data and preset operational boundary constraints, with the goal of maximizing comprehensive benefits, the target biomass co-firing ratio is calculated through the economic and environmental benefits optimization layer. Based on the target biomass blending ratio, with the goal of maximizing boiler efficiency, a nonlinear mapping relationship between operating parameters and combustion results is established and optimized through the combustion process optimization layer to generate the target operating parameters.

4. The intelligent biomass co-firing method for coal-fired power plants according to claim 3, characterized in that, The step of calculating the target biomass co-firing ratio based on the multi-source data and preset operational boundary constraints, with the objective function of maximizing comprehensive benefits, through the economic and environmental benefit optimization layer, includes: Obtain a feasible range for the blending ratio; Multiple candidate blending ratios are generated within the feasible range of the blending ratio; For each candidate blending ratio, the comprehensive economic benefit score corresponding to the candidate blending ratio is calculated based on the multi-source data and the predicted values ​​of boiler efficiency, pollutant emission concentration and furnace temperature corresponding to the candidate blending ratio calculated by the combustion process optimization layer. From all candidate biomass blending ratios that satisfy the operational boundary constraints, the candidate biomass blending ratio with the highest comprehensive economic benefit score is selected as the target biomass blending ratio.

5. A method for intelligent co-firing of biomass in a coal-fired power plant according to claim 3, characterized in that, The step of generating the target operating parameters, based on the target biomass blending ratio and with maximizing boiler efficiency as the objective function, involves establishing a nonlinear mapping relationship between operating parameters and combustion results through the combustion process optimization layer and performing optimization. This includes: Obtain the adjustable range of the operation parameters; Within the adjustable range of the operating parameters, the optimal combination of operating parameters is searched based on a pre-trained boiler efficiency prediction model and a derivative-free optimization algorithm to obtain the target operating parameters.

6. A method for intelligent co-firing of biomass in a coal-fired power plant according to claim 2, characterized in that, The biomass characteristic data includes fuel moisture data and fuel heavy metal content; the control of the combustion process of the coal-fired power plant through a feedforward-feedback composite control strategy based on the target biomass blending ratio and corresponding target operating parameters includes: The primary air parameters are calculated based on the fuel moisture data, and the opening of the hot air damper and cold air damper of the primary air system are adjusted according to the primary air parameters to stabilize the coal mill outlet temperature. The corrosion risk index is calculated based on the heavy metal content of the fuel. The upper limit of the blending ratio is determined based on the corrosion risk index. An air distribution adjustment strategy is generated based on the heavy metal content of the fuel, so as to control the combustion process of the coal-fired power plant according to the air distribution adjustment strategy and the upper limit of the blending ratio.

7. The intelligent biomass co-firing method for coal-fired power plants according to claim 1, characterized in that, The method of controlling the combustion process of the coal-fired power plant through a feedforward-feedback composite control strategy based on the target biomass blending ratio and corresponding target operating parameters further includes: Using the rotational speed of the biomass feeder and the rotational speed of the coal feeder as control variables, and the actual co-firing ratio obtained in real time as the controlled variable, a PID control algorithm is adopted to coordinate and control the rotational speed of the biomass feeder and the rotational speed of the coal feeder so that the actual co-firing ratio approaches the target biomass co-firing ratio. Using the opening degree of the secondary air dampers at each level as the control variable, and the real-time measured concentration of nitrogen oxides and carbon content of fly ash in the flue gas as the controlled variables, a multivariate model predictive control algorithm is adopted to dynamically adjust the opening degree of the secondary air dampers at each level so that the concentration of nitrogen oxides approaches the set value and the carbon content of fly ash approaches the set value of carbon content.

8. A method for intelligent co-firing of biomass in a coal-fired power plant according to claim 5, characterized in that, Also includes: Based on the multi-source data, obtain the actual values ​​of the preset operating indicators; The deviation between the actual value and the model prediction value of the preset operating index is calculated to obtain the deviation value, wherein the model prediction value is the predicted value of the preset operating index calculated by the pre-trained boiler efficiency prediction model. The deviation value is compared with the deviation threshold. When the deviation value continues to be greater than the deviation threshold, the parameters of the pre-trained boiler efficiency prediction model are self-calibrated.

9. A method for intelligent co-firing of biomass in a coal-fired power plant according to claim 2, characterized in that, Also includes: Obtain actual fuel sample analysis data and operational performance data for the current blending cycle; The actual fuel sample analysis data describes the physicochemical properties of the fuel actually used, and the operating effect data describes the combustion response of the boiler system to the fuel actually used. The actual fuel sample analysis data and the operational performance data are correlated, and the correlated data is entered into the pre-built fuel characteristic database.

10. A smart biomass co-firing device for a coal-fired power plant, characterized in that, include: The data acquisition module is used to acquire multi-source data for coal-fired power plants, including inherent equipment parameters, external command and demand data, fuel attribute parameters, and real-time operation data. The decision module is used to calculate the target biomass co-firing ratio and the corresponding target operating parameters based on the multi-source data and preset operating boundary constraints through a multi-objective dynamic optimization model. The execution module is used to control the combustion process of the coal-fired power plant according to the target biomass blending ratio and the corresponding target operating parameters through a feedforward-feedback composite control strategy.

Images