A biomass blending combustion feeding system and a collaborative control method for a coal-fired boiler using existing coal storage facilities

Abstract

The application discloses a kind of coal-fired boiler biomass blending combustion feeding system and collaborative control method using existing coal storage facilities, it is related to the automatic control technical field of coal-fired boiler biomass coupling blending combustion, the system includes: acquisition and analysis module acquires the time series data of multiple devices, analyzes fuel stratification information and receives adjustment parameter;Stratification proportioning calculation module accounts for blending ratio, generates basic feeding instruction;Disturbance complex correction module quantifies disturbance and corrects instruction;Variable load collaborative decision module issues fuel feeding instruction in combination with load and storage material;Closed-loop feedback optimization module iteratively updates various parameters according to boiler data and feeds back each module;The application adopts disturbance correction, variable load decision and closed-loop iteration multilayer regulation and control, quantifies various discharging disturbances and corrects feeding instruction, adjusts variable load feeding under working condition, relies on operating data to update parameters autonomously, adapts to working condition dynamic change, reduces manual setting, stabilizes combustion and controls pollutant emission.

Description

Technical Field

[0001] This invention relates to the field of automatic control technology for biomass co-firing in coal-fired boilers, specifically to a feeding system and collaborative control method for biomass co-firing in coal-fired boilers utilizing existing coal storage facilities. Background Technology

[0002] The dual-carbon development goals have driven the retrofitting of coal-fired power units with biomass. Relying on biomass briquettes to replace part of the raw coal can reduce fossil energy consumption and reduce carbon emissions throughout the unit's operation. This is the mainstream technology path for energy conservation and carbon reduction in the thermal power industry. Most coal-fired power plants have already built standardized coal storage silos and supporting coal feeders and other material storage and supply hardware facilities. If dedicated biomass storage and conveying equipment is added separately, it will significantly increase the project's construction investment and site occupation costs. The industry generally tends to modify existing coal storage facilities to achieve layered storage of the two fuels in a shared silo. In the shared coal storage silo operation mode, raw coal and biomass briquettes are stacked in layers and then fed into the boiler for combustion simultaneously. The fuel blending ratio directly determines the furnace combustion stability, steam parameter stability, and flue gas pollutant emission indicators. The current unit automation control system can simultaneously collect multiple types of operating data such as material level, conveying flow rate, unit load, furnace combustion, and steam heat. Digital monitoring hardware has the conditions for full-domain deployment, providing hardware support for building a multi-parameter linkage intelligent feeding control system. The industry urgently needs an integrated blending feeding control solution adapted to the shared coal storage silo scenario.

[0003] Existing biomass co-firing feeding control schemes are mostly designed for scenarios where raw coal and biomass are stored in separate bins. They cannot adapt to single-bin, layered stacking conditions, making it difficult to distinguish the two fuel layers and their real-time inventory levels. Inventory calculation relies solely on historical feed rate differences, which accumulates measurement deviations over time, causing significant discrepancies between the instantaneous blending ratio and the design value. Conventional control logic lacks quantitative identification and pre-compensation mechanisms for feeding disturbances such as material collapse, bridging, and low-level flow rate reduction within the bin. It relies solely on lagging adjustment of the feeder speed based on boiler back-end parameters, resulting in large fluctuations in combustion parameters after disturbances occur. This can easily lead to problems such as steam pressure instability, furnace oxygen content deviation, and instantaneous exceedances of pollutants. During load increases and decreases, the system does not differentiate between different load conditions and match different feeding distribution logics. The adjustment range is unconstrained, which can easily cause sudden changes in the coal feeder speed. Furthermore, control parameters rely on manual periodic tuning and cannot be dynamically and autonomously updated based on fuel properties, bin feeding status, and unit operating conditions, leading to a continuous decline in control accuracy over long-term operation. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a biomass blending and feeding system and collaborative control method for coal-fired boilers that utilizes existing coal storage facilities. This system achieves layered storage of raw coal and biomass briquettes based on existing coal storage silos, eliminating the need for additional dedicated storage equipment. The system comprises five functional modules. It collects multi-dimensional time-series data on silo level, feeding and conveying, unit load, and boiler combustion at high frequency. Combining silo geometry and fuel property parameters, it analyzes the layered interface and real-time inventory. A dedicated algorithm calculates the instantaneous blending ratio to generate basic feeding instructions. The system quantitatively identifies various feeding disturbances within the silo and completes dual-layer instruction correction. It classifies operating conditions based on unit load change characteristics to achieve collaborative feeding adjustment. Simultaneously, a closed-loop feedback link is established to continuously iterate and optimize various control parameters.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In one aspect, a biomass co-firing feeding system for a coal-fired boiler utilizing existing coal storage facilities, the system comprising: The data acquisition and analysis module connects the material level detection device, the coal feeder metering device, the boiler detection subsystem and the unit load control subsystem. It collects time-series data and analyzes the stratification interface, stacking height and real-time inventory of raw coal and biomass briquettes based on material density and bin geometric parameters. It also receives the initial input stratification identification and adjustment parameters or receives the stratification identification and adjustment parameters returned by the closed-loop feedback optimization module. The layered proportioning calculation module receives layered material level time-series data and periodic cumulative feeding quality data, uses the layered time-series blending proportioning algorithm to calculate the instantaneous blending ratio, calculates the difference between the instantaneous blending ratio and the preset proportion, generates a basic feeding coordination instruction based on the difference, and receives the layered mixing compensation coefficient. The disturbance composite correction module receives the basic feeding coordination command and the instantaneous blending ratio, uses the spatiotemporal coupled disturbance quantification algorithm to calculate the quantification value of the comprehensive characteristics of the feeding disturbance, corrects the basic command based on the quantification value and outputs the corrected command, and receives the disturbance feature weight coefficient. The variable load collaborative decision-making module receives the corrected feeding control command, retrieves the fuel inventory and unit load data in the bin, pre-stores the raw coal safe storage threshold and the biomass maximum stacking threshold, divides the variable load working conditions according to the load change rate and the real-time load interval and matches the increase and decrease allocation logic, generates and issues the raw coal feeder collaborative command and the biomass briquette fuel feeder collaborative command. The closed-loop feedback optimization module receives boiler operating parameters and the quantified values, updates the thresholds using a partitioned threshold adaptive iterative algorithm, generates adjustment parameters, compensation coefficients, and weighting coefficients, and feeds them back to the data acquisition and parsing module, the hierarchical proportioning calculation module, and the disturbance composite correction module.

[0006] Furthermore, when the data acquisition and parsing module performs layered parsing, it retrieves parameters including the total vertical height of the coal storage silo, the cross-sectional area of ​​the coal storage silo, the shrinkage dimension of the bottom discharge port of the coal storage silo, the bulk density per unit volume of raw coal, the bulk density per unit volume of biomass briquettes, the bulk density correction coefficient of raw coal, the bulk density correction coefficient of biomass briquettes, the fixed parameters of the dielectric constant of raw coal and biomass briquettes. The calculation object of the layered parsing is the three-dimensional point cloud value of the entire material surface in the silo output by the material level detection device. The output values ​​of the layered parsing include the vertical height of the layered interface, the spatial tilt angle of the layered interface, the real-time remaining total mass of raw coal, the real-time remaining total mass of biomass briquettes, the vertical stacking height of raw coal, the vertical stacking height of biomass briquettes, and the remaining effective volume of material that can be discharged in the silo.

[0007] Furthermore, the layered blending calculation module pre-inputs fixed parameters including the total vertical height of the coal storage silo, the cross-sectional area of ​​the coal storage silo, the shrinkage dimension of the bottom discharge port of the coal storage silo, the bulk density of raw coal per unit volume, the bulk density of biomass briquettes per unit volume, the bulk density correction coefficient of raw coal, the bulk density correction coefficient of biomass briquettes, and the target blending mass ratio of raw coal and biomass briquettes under the steady-state operation of the unit. The calculation steps for the layered blending calculation module to generate the basic feeding coordination command are as follows: read the instantaneous blending ratio value output by the layered time-series blending ratio solution algorithm, read the pre-input target blending mass ratio value, calculate the difference between the two sets of values, generate the layered time-series deviation compensation adjustment amount based on the difference, and superimpose the layered time-series deviation compensation adjustment amount onto the feeder reference speed control value. After superposition, the value is used as the basic feeding coordination command.

[0008] Furthermore, the disturbance composite correction module internally stores a material feeding disturbance feature library containing three types of disturbance conditions corresponding to time-series fluctuation value ranges. The first type is the material collapse condition, corresponding to the storage range of short-term incremental values ​​of instantaneous flow rate and short-term incremental values ​​of material level drop. The second type is the material bridging condition in the bin, corresponding to the storage range of the duration of the instantaneous flow rate continuously decreasing to near zero and the storage range of the duration of the material level height stopping decreasing. The third type is the low material level flow rate decay condition, corresponding to the storage range of the flow rate synchronous decay ratio when the height of the remaining material in the bin is less than one-fifth of the total bin height. The disturbance composite correction module uses a sliding window with a fixed duration of five seconds. The instantaneous flow rate time-series value of the coal feeder and the time-series value of the material level drop in the bin are continuously recorded in the window. All values ​​in the window are matched with the value ranges stored in the material feeding disturbance feature library. The matching operation output includes a disturbance condition classification identifier and a disturbance intensity classification identifier.

[0009] Furthermore, the disturbance composite correction module corrects the basic feeding coordination command in two steps. The first step reads the disturbance condition classification identifier and the disturbance intensity classification identifier, generates an adaptive feedforward compensation adjustment value based on the two sets of identifiers, and adds the adaptive feedforward compensation adjustment value to the basic feeding coordination command. The second step reads the furnace oxygen content value, steam pipeline pressure value, and pollutant emission concentration value output by the boiler detection subsystem, reads the steady-state operation target value corresponding to each of the three types of parameters, calculates the deviation between the real-time value and the target value of each set of parameters, and calculates the deviation rate per unit time between the two sets of deviation values. Based on the two sets of deviation values, a fine feedback correction adjustment value is generated, and the fine feedback correction adjustment value is added to the command value after the first step is completed. The value after the addition is used as the corrected feeding control command.

[0010] Furthermore, the variable load collaborative decision-making module pre-stores the raw coal safe storage threshold as the raw coal storage mass value corresponding to 15% of the total effective volume of the coal storage silo, and the variable load collaborative decision-making module pre-stores the maximum biomass storage threshold as the biomass briquette fuel storage mass value corresponding to 40% of the total effective volume of the coal storage silo. Before each generation of the coal feeder collaborative instruction, the variable load collaborative decision-making module performs a storage quantity verification operation, reads the real-time remaining total mass value of raw coal and the real-time remaining total mass value of biomass briquettes output by the data acquisition and analysis module, and performs a size comparison operation between the two sets of storage quantity values ​​and the two types of pre-stored threshold values.

[0011] Furthermore, the variable load collaborative decision-making module classifies the variable load condition judgment value as the load change rate of 2% of the rated unit load per minute. The variable load collaborative decision-making module divides the variable load condition judgment conditions into three categories: the first category is that the load increase rate per unit time is greater than 2% of the rated unit load per minute; the second category is that the load decrease rate per unit time is greater than 2% of the rated unit load per minute, and the real-time unit load value is less than 50% of the rated unit load; the third category is that the unit load value decreases by more than 30% of the rated unit load per unit time, and the load decrease rate is less than or equal to 2% of the rated unit load per minute. The variable load collaborative decision-making module executes the increase / decrease allocation logic in the order of the first, second, and third categories of conditions. The variable load collaborative decision-making module internally stores the feed adjustment rate limit value as the feeder speed adjustment amount per second does not exceed 5% of the rated speed. Before generating the raw coal feeder collaborative instruction and the biomass briquette feeder collaborative instruction, a limit verification calculation is performed to verify the relationship between the unit time instruction adjustment value and the limit value.

[0012] Furthermore, the closed-loop feedback optimization module receives boiler operating parameters including real-time calorific value of flue gas entering the furnace, boiler main steam pressure, boiler main steam temperature, oxygen content at the furnace outlet, real-time emission concentration of flue gas pollutants, steam parameter response time under unit load change conditions, and duration of boiler combustion parameter fluctuations after feeding disturbances. The closed-loop feedback optimization module performs deviation calculations on each set of boiler operating parameters, reads the intermediate value of the target steady-state operating range corresponding to the parameter, calculates the magnitude of the deviation between the real-time parameter value and the intermediate value, the number of sampling periods in which the deviation persists, and the increase or decrease of the deviation per unit time, and stores the deviation calculation results for all sampling periods.

[0013] Furthermore, after completing the threshold update, the closed-loop feedback optimization module decomposes and outputs four types of values: the first type is the stratified mixing deviation compensation coefficient, the second type is the flow fluctuation characteristic weight coefficient, the third type is the material level drop rate fluctuation weight coefficient, and the fourth type is the stratified blending ratio fluctuation coupling weight coefficient. The closed-loop feedback optimization module sets up three independent data transmission branches to complete parameter feedback. The first branch transmits the stratified mixing deviation compensation coefficient to the stratified blending ratio calculation module. The second branch transmits the flow fluctuation characteristic weight coefficient, the material level drop rate fluctuation weight coefficient, and the stratified blending ratio fluctuation coupling weight coefficient to the disturbance composite correction module. The third branch transmits the long-term cumulative boiler operating parameter deviation calculation result to the stratified blending ratio calculation module.

[0014] On the other hand, a collaborative control method for a biomass co-firing feeding system in a coal-fired boiler utilizing existing coal storage facilities, the specific steps of which are as follows: S1. The data acquisition and analysis module collects time-series data of material level, feed flow rate, unit load, and boiler flue gas and steam parameters at a frequency of 1Hz; it retrieves parameters of bin geometry, fuel density, and dielectric constant; and analyzes raw coal, biomass stratification interface, stacking height, and real-time inventory based on the three-dimensional material surface point cloud in the bin, and outputs stratified inventory data. S2. The layered proportioning calculation module reads the layered stock time-series data, uses the layered time-series blending ratio calculation algorithm to calculate the instantaneous blending ratio, compares it with the preset target blending ratio to obtain the deviation compensation amount, and adds it to the reference speed to generate the basic feeding coordination command. S3, the disturbance composite correction module matches three types of disturbance conditions—collapse, bridging, and low material level—through a 5-second sliding window, and calculates the comprehensive quantification value of the disturbance using a spatiotemporal coupled disturbance quantification algorithm; first, it performs feedforward compensation based on the disturbance condition, and then performs fine feedback correction based on furnace oxygen, steam pressure, and pollutant parameters, and outputs the corrected feeding control command. S4. The variable load collaborative decision-making module verifies whether the storage capacity of raw coal and biomass silos meets the storage threshold; it divides three types of variable load conditions according to the load rise and fall rate and matches the raw coal and biomass increase and decrease distribution logic; it verifies according to the limit rule that the speed adjustment per second does not exceed 5% of the rated speed and issues collaborative feeding instructions to the raw coal feeder and biomass feeder respectively. S5, the closed-loop feedback optimization module collects boiler operating parameters throughout the day and calculates the steady-state deviation of each parameter; it uses a partitioned threshold adaptive iterative algorithm to update the disturbance judgment threshold, and generates hierarchical identification adjustment parameters, mixing compensation coefficients, and disturbance weight coefficients, which are fed back to the data acquisition and analysis module, hierarchical proportioning calculation module, and disturbance composite correction module through three branches, and achieve full-process adaptive optimization through iterative loop.

[0015] Compared with existing technologies, this biomass co-firing feeding system and collaborative control method for coal-fired boilers utilizing existing coal storage facilities has the following advantages: I. This invention integrates multi-dimensional time-series data acquisition with a three-dimensional point cloud analysis architecture for the material surface within the storage silo. Relying on multi-source information linkage calculation of inventory, flow, and load, it achieves accurate identification of the stratification status and real-time inventory of raw coal and biomass briquettes. It eliminates the need to rely on historical feeding differences to verify inventory, significantly reducing the cumulative error in stratification ratio calculation. The system establishes a stratification ratio calculation link, dynamically generating basic feeding instructions based on the deviation between the real-time stratification blending ratio and the target ratio. Simultaneously, a stratification mixing compensation mechanism is provided to offset the ratio deviation caused by fuel mixing, continuously and stably maintaining the fuel blending ratio entering the furnace within the design range. The entire architecture reuses existing coal storage silos and supporting coal feeder hardware, eliminating the need for large-scale modification of coal storage facilities. Relying on software algorithms, it achieves coordinated metering and ratio control of stratified fuels, adapting to fuel storage conditions with different stacking heights and stacking sequences. It effectively solves the operational pain points of traditional blending systems, such as the inability to distinguish stratified materials within the silo, large ratio fluctuations, and inaccurate inventory calculations, thereby improving the basic stability of fuel blending control.

[0016] II. This invention employs a dual-layer control logic of disturbance composite correction and variable load collaborative decision-making, superimposed with a closed-loop adaptive iterative optimization link, to construct a full-process control system of feedforward compensation, feedback correction, and online parameter self-updating. It establishes a quantitative evaluation model for various material feeding disturbances such as material collapse, bridging, and low material level decay. A sliding window is used to identify the type and intensity of disturbances, and dual-layer command correction is completed simultaneously with boiler combustion flue gas, steam, and pollutant parameters, reducing the impact of material disturbances on combustion conditions. In variable load scenarios, operating conditions are distinguished based on load change characteristics, and adjustment limits are set, balancing unit load response speed and feeding equipment operation safety. Simultaneously, fuel overstocking and material shortages are avoided by relying on existing threshold constraints. The closed-loop module continuously collects boiler operating deviation data and iteratively updates various control coefficients and judgment thresholds, autonomously adapting to dynamic changes in fuel properties, unit operating conditions, and material feeding status throughout the entire process. This reduces the workload of manual parameter tuning and achieves long-term stable optimization of boiler combustion efficiency and pollutant emission levels.

[0017] Other advantages, objectives and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination or study, or may be learned from the practice of the invention. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 The flowchart shows the logic flow of each module in the biomass co-firing feeding system for coal-fired boilers that utilize existing coal storage facilities. Figure 2 A flowchart outlining the overall steps of a collaborative control method for a biomass co-firing feeding system in a coal-fired boiler utilizing existing coal storage facilities; Figure 3 This is a block diagram of the data input / output operation logic for the layered proportioning calculation module. Detailed Implementation

[0020] To further illustrate the technical means and effects adopted by the present invention to achieve the intended purpose, the following detailed description of the specific implementation methods, structures, features and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided.

[0021] Example: Reference Figure 1One embodiment of the present invention proposes a biomass co-firing feeding system and collaborative control method for coal-fired boilers utilizing existing coal storage facilities. It employs a data processing mechanism that includes layered material surface point cloud analysis, layered time-series blending ratio calculation, spatiotemporal coupling disturbance quantification correction, variable load condition tiered collaborative allocation, and zoned threshold adaptive iterative closed-loop optimization. This mechanism enables the layered storage and real-time precise feeding of raw coal and biomass briquettes without adding independent storage silos, relying solely on existing coal storage facilities. It suppresses disturbances caused by material collapse, bridging, and low-level flow attenuation within the silos that affect the blending ratio. It adapts to wide-range variable load operation of the unit, achieves self-iterative optimization of operating parameters throughout the process, and ensures stable and controllable boiler flue gas calorific value, steam parameters, and pollutant emissions.

[0022] The biomass co-firing feeding system for coal-fired boilers utilizing existing coal storage facilities described in this embodiment specifically includes: The data acquisition and analysis module connects to the material level detection device, coal feeder metering device, unit load control subsystem, and boiler detection subsystem at its signal input terminals. It synchronously acquires full-dimensional time-series data from the four types of devices at a fixed acquisition frequency of 1Hz. It locally stores static parameters such as the geometric dimensions of the coal storage silo, fuel bulk density, compaction correction coefficient, and fuel dielectric constant. Based on the three-dimensional point cloud of the material surface in the silo, it performs layered interface analysis and calculation of raw coal and biomass briquettes, and outputs layered storage data such as layered interface height, layered interface tilt angle, real-time raw coal inventory, real-time biomass briquettes inventory, and effective material discharge volume in the silo. The layered storage data is then transmitted to the layered proportioning calculation module. The layered blending ratio calculation module receives layered stock time-series data from the data acquisition and analysis module and layered mixing deviation compensation coefficients from the closed-loop feedback optimization module. The module has a built-in layered time-series blending ratio solution algorithm. The algorithm calculates the instantaneous blending ratio of raw coal and biomass briquettes at time t. The difference between the instantaneous blending ratio and the pre-stored steady-state target blending mass ratio is calculated to generate the layered time-series deviation compensation adjustment amount. The compensation adjustment amount is then superimposed on the feeder's reference speed control value to generate a basic feeding coordination command and transmit it to the disturbance composite correction module. The disturbance composite correction module receives inputs from the basic feeding coordination command output by the layered proportioning calculation module, the instantaneous blending ratio output by the layered time-series blending ratio calculation module, and the boiler combustion parameters transmitted by the boiler detection subsystem. The module includes a built-in fixed sliding window with a duration of 5 seconds and a feeding disturbance feature library. The sliding window continuously records the instantaneous flow rate of the coal feeder and the time-series data of the material level drop in the silo. It then performs matching calculations between the time-series data within the window and the feeding disturbance feature library to identify three types of disturbance conditions: material collapse, bridging, and low-level flow attenuation, and outputs disturbance condition classification labels and disturbance intensity grading labels. The module is equipped with a spatiotemporal coupling... The module employs a disturbance quantification algorithm to calculate the comprehensive characteristic quantification value of the feeding disturbance. The module completes the feeding command correction calculation in two steps: First, it generates an adaptive feedforward compensation adjustment amount based on the disturbance condition classification and disturbance intensity grading, which is then superimposed on the basic feeding coordination command. Second, it reads the real-time values ​​and steady-state target values ​​of furnace oxygen content, steam pipeline pressure, and pollutant emission concentration, calculates the magnitude of the deviation and the rate of change of the deviation per unit time for each set of values, generates a fine feedback correction adjustment amount, superimposes it on the feedforward correction command, outputs the corrected feeding control command, and transmits it to the variable load collaborative decision-making module. The variable load collaborative decision-making module receives, at its input end, the corrected feeding control command issued by the disturbance composite correction module, the real-time inventory of raw coal and biomass output by the data acquisition and analysis module, and the unit load time sequence data output by the unit load control subsystem. The module locally stores the safe storage threshold of raw coal, the maximum storage threshold of biomass, the variable load condition judgment benchmark, and the feed speed adjustment limit value. The module sequentially executes the inventory verification calculation, the variable load condition judgment calculation, and the speed adjustment limit verification calculation, and matches the raw coal and biomass fuel increase and decrease allocation logic according to the operating condition priority. Finally, it generates the raw coal feeder collaborative command and the biomass briquette fuel feeder collaborative command respectively, and issues them to the two types of feeder actuators. The closed-loop feedback optimization module receives all-time boiler operating parameters from the boiler detection subsystem and the quantitative value of the comprehensive characteristics of material feeding disturbance from the disturbance composite correction module. The module performs deviation calculations on each set of boiler operating parameters, reads the intermediate value of the steady-state target range of the parameters, and calculates the magnitude of the deviation between the real-time parameters and the intermediate value, the number of continuous sampling periods for the deviation, and the increase or decrease in deviation per unit time. It locally stores the deviation calculation results for all sampling periods. The module is equipped with a partitioned threshold adaptive iterative algorithm to update the material feeding disturbance partition judgment threshold. After the threshold update, it outputs the layered mixing deviation compensation coefficient, flow fluctuation characteristic weight coefficient, material level descent rate fluctuation weight coefficient, and layered blending ratio fluctuation coupling weight coefficient. The module has three independent data transmission branches: the first branch transmits the layered mixing deviation compensation coefficient back to the layered proportion calculation module; the second branch transmits the three types of disturbance weight coefficients back to the disturbance composite correction module; and the third branch transmits the long-term cumulative boiler operating parameter deviation calculation results back to the layered proportion calculation module, iteratively completing the full-process adaptive parameter optimization.

[0023] Specifically, the implementation logic of this invention lies in building a five-layer cascaded, closed-loop self-correcting layered co-firing and feeding control architecture. This architecture first relies on a data acquisition and analysis module to accurately perceive the inventory and material layer interface of the two types of layered fuels within the coal storage silo, eliminating measurement errors caused by silo geometry and fuel properties. Then, a layered proportioning calculation module dynamically calculates the instantaneous blending ratio based on real-time layered inventory, eliminating deviations between static proportions and actual feeding conditions. A disturbance composite correction module establishes a feedforward + feedback dual-layer correction link to address inherent feeding disturbances in the coal storage silo, simultaneously suppressing proportion deviations caused by material disturbances within the silo and fluctuations in boiler combustion parameters. A variable load collaborative decision-making module combines silo inventory constraints with dynamic changes in unit load, allocating the feeding adjustment amounts of the two types of fuels in stages and limiting the adjustment rate to avoid material shortages and over-proportioning during significant load fluctuations. Finally, a closed-loop feedback optimization module collects long-term boiler combustion operation data, iteratively updating the compensation, weight, and threshold parameters required for the entire chain calculation, forming a complete closed-loop control link from silo material layer perception, feeding command generation, disturbance correction, variable load allocation, to parameter self-optimization. The first four modules complete real-time open-loop control calculations, while the fifth module completes long-cycle parameter iterative optimization. The two modules operate in parallel, balancing instantaneous feeding control accuracy with long-term adaptive capability.

[0024] Optionally, the time-series data collected by the data acquisition and parsing module includes the time-series values ​​of material level height, instantaneous conveying flow rate of the coal feeder and the cumulative material feeding mass in a single cycle, the real-time load of the unit and the rate of change of load per unit time, the calorific value of flue gas, the concentration of pollutant emissions, the oxygen content in the furnace, the negative pressure in the furnace, the pressure of the steam pipeline, and the temperature of the steam pipeline. All signals are synchronously acquired at a unified acquisition frequency of once per second.

[0025] Specifically, the entire data acquisition chain is configured with a unified system master clock, synchronously triggering analog-to-digital conversion of all detection devices to ensure complete alignment of timestamps for the output timing data of the four types of equipment: material level, feed, unit, and boiler, avoiding timing misalignment errors caused by asynchronous sampling of multiple devices. The material level detection device outputs a full-domain three-dimensional point cloud of the material surface within the silo, including the material level height information corresponding to all coordinate points within the silo; the coal feeder metering device refreshes the instantaneous conveying flow rate every second and accumulates the total mass of material fed within each whole second cycle; the unit load control subsystem synchronously outputs the current actual load of the unit and the load increase / decrease rate per minute; the boiler detection subsystem synchronously collects combustion parameters of flue gas, furnace, and steam in all dimensions. For example, under 1Hz synchronous acquisition conditions, a complete set of time-series data is captured synchronously at a certain sampling moment: the material surface point cloud includes 2000 sets of three-dimensional coordinate material level height values ​​within the bin; the instantaneous flow rate of the raw coal feeder is 12.6 t / h, with a cumulative discharge of 3.5 kg per cycle; the instantaneous flow rate of the biomass feeder is 3.4 t / h, with a cumulative discharge of 0.95 kg per cycle; the unit's real-time load is 60% of the rated load, with a load change rate of +0.8% of the rated load / minute; the lower heating value of the flue gas is 16.2 MJ / kg, the NOx emission concentration is 68 mg / m³, the oxygen content at the furnace outlet is 3.2%, the furnace negative pressure is -85 Pa, the main steam pressure is 13.7 MPa, and the main steam temperature is 542℃. All data are accompanied by a unified timestamp and are synchronously transmitted to the data acquisition and parsing module for hierarchical parsing calculations.

[0026] Optionally, the static parameters retrieved by the data acquisition and parsing module when performing layered parsing include the total vertical height of the coal storage silo, the cross-sectional area of ​​the silo, the shrinkage size of the bottom discharge port, the bulk density per unit volume of raw coal and biomass briquettes, the packing density correction coefficients of the two types of fuels, and the fixed parameters of the dielectric constants of raw coal and biomass. The calculation object of the layered parsing is the three-dimensional point cloud value of the material surface in the entire silo. The layered parsing output includes the vertical height of the layered interface, the spatial tilt angle of the layered interface, the real-time remaining total mass of raw coal, the real-time remaining total mass of biomass, the vertical stacking height of raw coal, the vertical stacking height of biomass, and the remaining effective volume of material that can be discharged in the silo.

[0027] Specifically, the layered analysis calculation consists of two steps. The first step distinguishes between raw coal and biomass briquettes within the silo based on the 3D material surface point cloud and fuel dielectric constant parameters. It fits the interface between the two fuel types, calculates the vertical height and spatial tilt angle of the layered interface, and eliminates measurement deviations in material level caused by localized material accumulation tilt. The second step combines silo geometric parameters, fuel bulk density, and compaction correction coefficients to calculate the current effective bulk volume of raw coal and biomass fuel respectively. This is then converted to obtain the real-time remaining total mass of the two fuel types, while simultaneously deducting the ineffective volume of the discharge port contraction area, outputting the remaining effective volume available for material discharge within the silo. During the calculation, two options can be chosen to complete the inventory calculation: one is to directly read the real-time total mass value obtained from the material surface point cloud analysis, skipping the volume density conversion; the other is to deduce the effective volume through the material layer height and silo geometric function, and calculate the inventory based on density parameters. Throughout the process, the difference between the current inventory and the historical cumulative discharge mass is not used to infer the inventory, avoiding the continuous accumulation of measurement errors. For example, the static preset parameters of a coal storage silo are as follows: total silo height 12m, silo cross-sectional area 28㎡, bottom discharge port constriction section height 1.2m; raw coal bulk density 0.92t / m³, compaction correction factor 0.96, dielectric constant 3.8; biomass briquetted fuel bulk density 0.38t / m³, compaction correction factor 0.91, dielectric constant 2.1. Importing real-time 3D material surface point cloud data to complete layered analysis, the fitting yielded a layered interface vertical height of 7.3m and an interface tilt angle of 3.6°; the vertical stacking height of raw coal was 7.3m, and the vertical stacking height of biomass was 3.5m; after deducting the ineffective volume from bottom shrinkage, the effective stacking volume of raw coal was 182.4m³, and the effective stacking volume of biomass was 34.6m³; the real-time remaining total mass of raw coal was calculated to be approximately 161.2t (182.4×0.92×0.96), and the real-time remaining total mass of biomass was approximately 11.9t (34.6×0.38×0.91). The remaining effective volume of material that can be fed into the silo was 217.0m³. All layered inventory data were output to the layered proportioning calculation module.

[0028] Optionally, the stratified blending calculation module pre-inputs static parameters including the geometric dimensions of the coal storage bin, the bulk density and compaction correction coefficient of the two types of fuels, and the target blending mass ratio of raw coal and biomass during steady-state operation of the unit. The calculation steps for the stratified blending calculation module to generate the basic feeding coordination command are as follows: reading the instantaneous blending ratio output by the stratified time-series blending ratio solution algorithm, reading the preset target blending ratio, calculating the difference between the two sets of values, generating a stratified time-series deviation compensation adjustment amount, superimposing the compensation adjustment amount onto the feeder's reference speed control value, and using the superimposed value as the basic feeding coordination command. The mathematical expression corresponding to the stratified time-series blending ratio solution algorithm is: In the formula, The ratio of raw coal to biomass briquettes at instantaneous outflow mass at time t; Represents the bulk density per unit volume of biomass briquettes; Represents the bulk density per unit volume of raw coal; This represents the remaining vertical stacking height of biomass briquettes in the silo at time t. This represents the remaining vertical stacking height of raw coal in the warehouse at time t; and These are, respectively, the effective stacking volume functions of biomass briquettes and raw coal corresponding to the current vertical stacking height, derived from fixed geometric parameters such as the total vertical height of the coal storage silo, cross-sectional area, and shrinkage size of the bottom discharge port; The stratification interface mixing deviation compensation coefficient represents the coefficient of error at time t. pass Calculated, in the formula This is the compensation ratio constant for stratified mixing. This represents the number of historical sampling periods. For the front The instantaneous mixing ratio for each sampling period, For the steady-state target blending mass ratio, in this formula... A fixed constant is preset manually to adjust the correction range of the compensation coefficient. The summation term represents the total number of sampling points within the sliding statistical window. The summation term is the average of the absolute deviations between the historical instantaneous blending ratios and the target ratio within the window. This overall representation characterizes the degree of ratio deviation caused by long-term stratified blending, thereby quantifying the real-time stratified blending deviation compensation coefficient. .

[0029] Specifically, the stratified time-series blending ratio solution algorithm calculates the effective volume of the two types of fuel based on the real-time material layer height and bin geometry function output by stratified analysis. It then combines this with the bulk density to obtain the real-time stock ratio, and finally superimposes the stratified mixing deviation compensation coefficient issued by the closed-loop feedback module to obtain the true instantaneous blending ratio. The instantaneous blending ratio is then compared with the preset steady-state target ratio. A positive difference indicates a low biomass feed ratio, requiring an increase in the biomass feeder's base speed; a negative difference indicates a high biomass feed ratio, requiring a decrease in the biomass feeder's base speed. The compensation adjustment is linearly superimposed to the base speed, generating a basic feeding coordination command, which is then sent to the disturbance composite correction module. Figure 3 As shown.

[0030] For example, the unit's steady-state preset target blending ratio of raw coal and biomass is... Hierarchical analysis output at time t , Substitute into the warehouse body Geometric functions are obtained , , Closed-loop feedback sends out the current stratified mixing deviation compensation coefficient. Substitute into the formula to calculate: Instantaneous mass ratio of raw coal to biomass Compared with the target ratio of 4.0, the ratio difference is -0.02, generating a positive stratified timing deviation compensation adjustment amount, which is superimposed on the reference speed of the biomass feeder to obtain the basic feeding coordination command.

[0031] Optionally, the disturbance composite correction module internally stores a disturbance feature library containing three types of disturbance conditions and their time-series fluctuation ranges: the first type is material collapse, storing the short-term increment range of instantaneous flow rate and the short-term increment range of material level drop; the second type is material bridging in the bin, storing the duration range of instantaneous flow rate continuously approaching zero and the duration range of material level height stopping decline; the third type is low material level flow rate decay, storing the range of flow rate synchronous decay ratio when the material height in the bin is less than 1 / 5 of the total bin height. The disturbance composite correction module uses a 5-second fixed-duration sliding window to continuously input the time-series values ​​of the instantaneous flow rate of the coal feeder and the material level drop in the bin, matches all the time-series data in the window with the disturbance feature library ranges, and outputs a disturbance condition classification identifier and a disturbance intensity classification identifier.

[0032] The mathematical expression corresponding to the spatiotemporal coupling perturbation quantization algorithm is: In the formula, This represents the quantified value of the comprehensive characteristics of the disturbance caused by the material unloading from the coal storage bunker at time t. This represents the instantaneous fuel delivery flow rate of the coal feeder at time t; This represents the reference flow rate of the coal feeder in steady-state operation; This represents the decrease in material level within the silo during a single sampling period; This represents the standard material level drop during steady-state feeding. This refers to the instantaneous mixing ratio; The target blending ratio; Weighting coefficients for traffic fluctuation characteristics; Weighting coefficient for fluctuation in material level descent rate; Layered blending ratio fluctuation coupling weighting coefficient; flow fluctuation characteristic weighting coefficient Weighting coefficient for fluctuation in material level descent rate Layered blending ratio fluctuation coupling weighting coefficient The calculation formula is: , , In the formula , , The initial baseline values ​​for the three categories of weights are fixed. To determine the number of sliding window sampling periods for boiler parameters, , , They are respectively the front Real-time values ​​of main steam pressure, furnace oxygen content, and pollutant emission concentration during the cycle. , , These are the steady-state target values ​​for main steam pressure, furnace oxygen content, and pollutant emission concentration, respectively. The initial benchmark values ​​in the three sets of formulas are fixed offline calibration constants, and the summation mean term is the historical average deviation of the corresponding boiler combustion parameters. The greater the fluctuation of the combustion parameters, the greater the corresponding weight coefficient value. This dynamically quantifies the contribution ratio of the three types of disturbance factors to the overall feeding disturbance.

[0033] Specifically, a 5-second sliding window updates a set of flow rate and material level drop data every second, storing a total of 5 sets of time-series samples. It simultaneously traverses three types of disturbance characteristic intervals to complete matching and judgment, marking the current disturbance type and intensity level. Simultaneously, it uses a spatiotemporal coupled disturbance quantification algorithm to calculate the comprehensive disturbance quantification value; a larger quantification value indicates a stronger interference between the material feeding disturbance in the bin and the blending ratio. The correction of the basic feeding coordination command involves two steps: First, based on the disturbance classification and intensity, an adaptive feedforward compensation adjustment is generated and superimposed on the basic command to preemptively offset sudden flow changes caused by material collapse, bridging, and low material levels. Second, the real-time values ​​and steady-state target values ​​of furnace oxygen content, main steam pressure, and pollutant emission concentration are read, and the deviation between the real-time and target values ​​of each set of parameters and the rate of change of the deviation per unit time are calculated. Based on the two sets of deviations, a fine feedback correction adjustment is generated and superimposed on the feedforward correction command, outputting the final corrected feeding control command. Feedforward correction is used to eliminate instantaneous changes in the mix ratio caused by material disturbances in the silo, while feedback correction is used to offset the mix ratio deviation caused by lag fluctuations in boiler combustion parameters. The combination of the two achieves full-dimensional disturbance correction.

[0034] For example, steady-state reference flow steady-state standard material level drop Seconds; Current Instantaneous flow rate Single-cycle material level decrease Instantaneous blending ratio Target ratio Closed-loop feedback distributes weighting coefficients. , , Substitute into the perturbation quantization formula to calculate: Comprehensive perturbation quantization value The sliding window timing data matching determines a slight material collapse disturbance, generating a corresponding feedforward compensation adjustment amount that is superimposed on the basic command; the target oxygen content in the furnace is read simultaneously. ,real time Main steam pressure target ,real time NOx emission targets ,real time The system calculates the deviations and rates of change of the three sets of parameters, generates a fine feedback correction adjustment amount that is superimposed twice, and outputs the corrected feeding control command.

[0035] Optionally, the variable load collaborative decision-making module pre-stores the raw coal safe storage threshold as the raw coal storage mass corresponding to 15% of the total effective volume of the coal storage silo, and the biomass maximum stacking threshold as the biomass storage mass corresponding to 40% of the total effective volume of the coal storage silo; before each generation of the coal feeder collaborative instruction, a stock check is performed, the real-time remaining total mass of raw coal and biomass is read, and a comparison calculation is performed with the two types of thresholds respectively. The benchmark value for determining variable load conditions is the load change rate of 2% of the rated unit load per minute. Three categories of variable load conditions are defined: Category 1: Load increase rate per unit time > 2% / min of rated unit load; Category 2: Load decrease rate per unit time > 2% / min of rated unit load, and the real-time unit load < 50% of the rated unit load; Category 3: Load decrease rate per unit time > 30% of the rated unit load, and the load decrease rate ≤ 2% / min of the rated unit load. The variable load collaborative decision-making module executes the fuel increase / decrease allocation logic in the order of Category 1, Category 2, and Category 3 conditions. The internally stored feed adjustment rate limit value is that the feeder speed adjustment per second does not exceed 5% of the rated speed. Before generating the raw coal and biomass feeder collaborative instructions, a limit verification is performed, comparing the unit time instruction adjustment range with the limit value to avoid severe material feeding disturbances caused by large speed adjustments in a short period.

[0036] Specifically, the inventory verification logic is as follows: If the real-time remaining mass of raw coal is lower than the safe storage threshold, the adjustment range of the raw coal feeder is limited to prioritize ensuring the basic feed volume of raw coal; if the real-time remaining mass of biomass is higher than the maximum storage threshold, the continuous feeding of biomass is limited to avoid excessive accumulation in the silo. The variable load allocation logic is determined sequentially according to the priority of the working conditions, and only a single set of fuel increase / decrease adjustment rules is matched at the same time; the speed limit verification independently constrains the commands of the raw coal and biomass feeders. When the speed adjustment range in a single second exceeds 5% of the rated speed, the adjustment amount is forcibly cut off to the maximum limit value, and the feed flow is adjusted smoothly. For example, the total effective volume of the coal storage silo is 217m³, the volume corresponding to the safe storage threshold of raw coal is 15%, which is 32.55m³, equivalent to a raw coal mass of 32.55×0.92×0.96≈28.8t; the volume corresponding to the maximum storage threshold of biomass is 40%, which is 86.8m³, equivalent to a biomass mass of 86.8×0.38×0.91≈29.9t. The current raw coal inventory is 161.2t > 28.8t, and the biomass inventory is 11.9t < 29.9t. The inventory verification is passed and there is no storage constraint. The unit's rated load is 300MW, the current real-time load is 210MW, and the load increase rate is 2.6% of the rated load / minute, which meets the first type of load increase condition judgment condition and matches the load increase fuel synchronous incremental distribution logic. The calculated adjustment range of the raw coal feeder speed per second is 3.2% of the rated speed, and the adjustment range of biomass is 2.8% of the rated speed, both of which are less than the 5% limit threshold. The limit verification is passed, and two sets of coordinated feeding instructions are directly issued to the corresponding feeder actuators.

[0037] Optionally, the closed-loop feedback optimization module receives boiler operating parameters including real-time calorific value of flue gas entering the furnace, boiler main steam pressure, main steam temperature, oxygen content at the furnace outlet, real-time emission concentration of pollutants in flue gas, steam parameter response time under unit load change conditions, and duration of combustion parameter fluctuations after feeding disturbances. It performs deviation calculations on each set of boiler operating parameters, reads the intermediate value of the steady-state target interval of the parameters, calculates the magnitude of the deviation between the real-time parameters and the intermediate value, the number of consecutive sampling periods for the deviation, and the increase or decrease in the deviation per unit time, and stores the deviation calculation results for all sampling periods. The mathematical expression corresponding to the partition threshold adaptive iterative algorithm is: In the formula, represent Threshold for determining the disturbance zone during material feeding; Represents the iteration update Threshold for determining the disturbance zone during material feeding; The threshold iterative learning rate is represented by a fixed coefficient. The output of the spatiotemporal coupled disturbance quantization algorithm is the quantized value of the comprehensive characteristics of the material feeding disturbance. represent Real-time operating value of boiler main steam pressure; This represents the target value for the steady-state main steam pressure. Represents the constraint coefficient for deviation of boiler combustion parameters; The saturation limiting function is defined as: when the input value Less than the preset lower limit value Time to take ;when Greater than the preset upper limit value Time to take ; When it is in between, take itself; , For a preset positive constant value, satisfying To avoid negative correction factors and iteration divergence caused by large fluctuations in boiler pressure, and to ensure the stability of the threshold iteration direction; saturation limiting function input middle This is a pre-set combustion deviation constraint constant used to adjust the suppression strength of steam pressure deviation on the iterative correction term. The absolute deviation between the real-time value and the target value of the main steam pressure. Main steam pressure steady-state setpoint, saturation limiting function pass The calculation yields the following result: To perform the maximum value operation, To perform the minimum value operation, The lower limit of saturation The upper limit of saturation is used to constrain the iterative correction factor within the stable interval, preventing iterative calculations from diverging. The threshold iterative learning rate is a fixed coefficient. This is an offline calibration constant that controls the update step size of the perturbation judgment threshold during each iteration. This is the quantified value of the comprehensive characteristics of the material feeding disturbance obtained in real time. The threshold for perturbation partitioning is retained from the previous cycle. The difference between the two represents the degree of deviation between the current perturbation intensity and the original threshold. Overall, the perturbation threshold is adaptively and iteratively updated.

[0038] Specifically, the closed-loop feedback optimization module reads the boiler's full-dimensional operating parameters at fixed intervals, calculates the steady-state deviation for each parameter, and stores the time-series deviation dataset for a long time. It then substitutes these parameters into the adaptive iterative algorithm for partitioned thresholds to update the feeding disturbance partition judgment threshold used by the disturbance composite correction module. After the threshold update, it outputs four types of optimization parameters: stratified mixing deviation compensation coefficient, flow fluctuation characteristic weight coefficient, material level descent rate fluctuation weight coefficient, and stratified blending ratio fluctuation coupling weight coefficient. Three independent data transmission branches are set up to complete the parameter closed-loop feedback: the first branch transmits the stratified mixing deviation compensation coefficient to the stratified proportion calculation module; the second branch transmits the three types of disturbance weight coefficients to the disturbance composite correction module; and the third branch transmits the long-term cumulative boiler operating parameter deviation calculation results to the stratified proportion calculation module. This continuously iterates and optimizes the entire chain of calculation parameters to adapt to scenarios such as long-term unit operating condition drift, small changes in fuel properties, and attenuation of in-bin feeding characteristics.

[0039] For example, the iterative algorithm presets parameters such as learning rate. Combustion deviation constraint coefficient saturation lower limit Saturation upper limit The threshold for determining the original disturbance at time t. Current perturbation quantization value Actual value of main steam pressure Target pressure First, calculate the input terms of the saturation function: Between Between, the output of the SAT function Substitute the values ​​into the iterative formula to update the threshold: Updated perturbation detection threshold Simultaneous disassembly generates a new stratified mixing deviation compensation coefficient. Update weight coefficients , , The parameters are simultaneously distributed to the hierarchical ratio calculation module and the disturbance composite correction module through three branches to complete the single-cycle closed-loop parameter iterative optimization.

[0040] Reference Figure 2 This embodiment also provides a method for coordinated control of biomass co-firing feeding in coal-fired boilers using existing coal storage facilities. This method relies on the aforementioned co-firing feeding system, and the specific steps of the method are as follows: S1. The system's data acquisition and analysis module synchronously acquires full-dimensional time-series data from the material level detection device, coal feeder metering device, unit load control subsystem, and boiler detection subsystem at a fixed acquisition frequency of 1Hz; retrieves static parameters such as the geometric dimensions of the coal storage silo, fuel bulk density, compaction correction coefficient, and fuel dielectric constant; and completes the layer interface analysis of raw coal and biomass briquettes based on the three-dimensional point cloud of the material surface in the silo, outputting the layer interface height, tilt angle, real-time inventory of the two types of fuel, and effective material discharge capacity layer inventory data. S2. The system's layered blending ratio calculation module reads the layered stock time-series data, substitutes it into the layered time-series blending ratio calculation algorithm to calculate the instantaneous blending ratio at time t, performs a difference calculation between the instantaneous blending ratio and the preset steady-state target blending mass ratio, generates a layered time-series deviation compensation adjustment amount, adds it to the coal feeder's reference speed, and outputs the basic feeding coordination command. S3. The system's disturbance composite correction module continuously records the instantaneous flow rate and material level drop time series data through a fixed sliding window with a duration of 5 seconds. It matches and identifies three types of disturbance conditions—collapse, bridging, and low material level flow rate decay—with the built-in material feeding disturbance feature library. Simultaneously, it uses a spatiotemporal coupled disturbance quantification algorithm to calculate the comprehensive feature quantification value of the material feeding disturbance. The command correction is completed in two steps. The first step is to superimpose the adaptive feedforward compensation adjustment amount based on the disturbance condition classification label and the disturbance intensity classification label. The second step is to read the deviation and deviation change rate of the real-time values ​​of furnace oxygen content, steam pipeline pressure, and pollutant emission concentration from the steady-state target values, generate a fine feedback correction adjustment amount, superimpose it on the feedforward correction command, and output the corrected feeding control command. S4. The system's variable load collaborative decision-making module first retrieves the real-time inventory of raw coal and biomass, as well as the pre-stored safe storage threshold and maximum stacking threshold to complete the inventory verification. Based on the unit's load change rate, it divides the load into three types of variable load conditions, matches the raw coal and biomass fuel increase / decrease allocation logic according to the condition priority, sets a limit rule that the adjustment range of the coal feeder speed per second does not exceed 5% of the rated speed to complete the instruction verification, and issues independent collaborative feeding instructions to the raw coal feeder and the biomass briquette fuel feeder respectively. S5. The closed-loop feedback optimization module of the system continuously collects boiler operating parameters throughout the entire period, calculates the deviation of each set of parameters relative to the median value of the steady-state target range, the continuous sampling period of the deviation, and the rate of increase or decrease of the deviation, and stores the deviation calculation results of all cycles; it uses a partition threshold adaptive iterative algorithm to update the material feeding disturbance partition judgment threshold, decomposes and outputs the layered mixing deviation compensation coefficient and the three types of disturbance weight coefficients, and sends the optimization parameters back to the corresponding preceding calculation module through three independent data branches, and iteratively completes the full-process adaptive optimization.

[0041] It should be noted that the data transmission connections between the above modules do not necessarily mean direct point-to-point hardware connections. Indirect data transmission methods such as industrial Ethernet, 485 bus, and OPC communication can be used as long as the data interaction and control operation objectives of this invention are achieved.

[0042] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A biomass blending feeding system for a coal-fired boiler using existing coal storage facilities, characterized in that, The system includes: The data acquisition and analysis module connects the material level detection device, the coal feeder metering device, the boiler detection subsystem and the unit load control subsystem. It collects time-series data and analyzes the stratification interface, stacking height and real-time inventory of raw coal and biomass briquettes based on material density and bin geometric parameters. It also receives the initial input stratification identification and adjustment parameters or receives the stratification identification and adjustment parameters returned by the closed-loop feedback optimization module. The layered proportioning calculation module receives layered material level time-series data and periodic cumulative feeding quality data, uses the layered time-series blending proportioning algorithm to calculate the instantaneous blending ratio, calculates the difference between the instantaneous blending ratio and the preset proportion, generates a basic feeding coordination instruction based on the difference, and receives the layered mixing compensation coefficient. The disturbance composite correction module receives the basic feeding coordination command and the instantaneous blending ratio, uses the spatiotemporal coupled disturbance quantification algorithm to calculate the quantification value of the comprehensive characteristics of the feeding disturbance, corrects the basic command based on the quantification value and outputs the corrected command, and receives the disturbance feature weight coefficient. The variable load collaborative decision-making module receives the corrected feeding control command, retrieves the fuel inventory and unit load data in the bin, pre-stores the raw coal safe storage threshold and the biomass maximum stacking threshold, divides the variable load working conditions according to the load change rate and the real-time load interval and matches the increase and decrease allocation logic, generates and issues the raw coal feeder collaborative command and the biomass briquette fuel feeder collaborative command. The closed-loop feedback optimization module receives boiler operating parameters and the quantified values, updates the thresholds using a partitioned threshold adaptive iterative algorithm, generates adjustment parameters, compensation coefficients, and weighting coefficients, and feeds them back to the data acquisition and parsing module, the hierarchical proportioning calculation module, and the disturbance composite correction module.

2. The biomass blending feeding system for coal-fired boilers using existing coal storage facilities according to claim 1, characterized in that, When the data acquisition and analysis module performs layered analysis, it retrieves parameters including the total vertical height of the coal storage silo, the cross-sectional area of ​​the coal storage silo, the shrinkage dimension of the bottom discharge port of the coal storage silo, the bulk density per unit volume of raw coal, the bulk density per unit volume of biomass briquettes, the bulk density correction coefficient of raw coal, the bulk density correction coefficient of biomass briquettes, the fixed parameters of the dielectric constant of raw coal and biomass briquettes. The calculation object of the layered analysis is the three-dimensional point cloud value of the entire material surface in the silo output by the material level detection device. The output values ​​of the layered analysis include the vertical height of the layer interface, the spatial tilt angle of the layer interface, the real-time remaining total mass of raw coal, the real-time remaining total mass of biomass briquettes, the vertical stacking height of raw coal, the vertical stacking height of biomass briquettes, and the remaining effective volume of material that can be discharged in the silo.

3. A biomass co-firing feeding system for a coal-fired boiler utilizing existing coal storage facilities, as described in claim 1, is characterized in that... The pre-entered fixed parameters in the layered blending calculation module include the total vertical height of the coal storage silo, the cross-sectional area of ​​the coal storage silo, the shrinkage dimension of the bottom discharge port of the coal storage silo, the bulk density of raw coal, the bulk density of biomass briquettes, the bulk density correction coefficient of raw coal, the bulk density correction coefficient of biomass briquettes, and the target blending mass ratio of raw coal to biomass briquettes under the steady-state operation of the unit. The calculation steps for the layered blending calculation module to generate the basic feeding coordination command are as follows: read the instantaneous blending ratio value output by the layered time-series blending ratio solution algorithm, read the pre-entered target blending mass ratio value, calculate the difference between the two sets of values, generate the layered time-series deviation compensation adjustment amount based on the difference, and superimpose the layered time-series deviation compensation adjustment amount onto the feeder reference speed control value. After superposition, the value is used as the basic feeding coordination command.

4. A biomass co-firing feeding system for a coal-fired boiler utilizing existing coal storage facilities as described in claim 1, characterized in that, The disturbance composite correction module internally stores a material feeding disturbance feature library containing three types of disturbance conditions corresponding to time-series fluctuation value ranges. The first type is the material collapse condition, which corresponds to the storage of the short-term increment value range of instantaneous flow rate and the short-term increment value range of material level. The second type is the material bridging condition in the silo, which corresponds to the storage of the value range of the duration of the instantaneous flow rate continuously decreasing to near zero and the value range of the duration of the material level height value stopping decreasing. The third category is the low material level flow rate decay condition, which corresponds to the range of synchronous flow rate decay ratios when the height of the remaining material in the storage bin is less than one-fifth of the total height of the bin. The disturbance composite correction module uses a sliding window with a fixed duration of five seconds. The instantaneous flow rate of the coal feeder and the material level drop in the bin are continuously recorded inside the window. All values ​​in the window are matched with the value range stored in the material feeding disturbance feature library. The matching operation output includes the disturbance condition classification label and the disturbance intensity classification label.

5. A biomass co-firing feeding system for a coal-fired boiler utilizing existing coal storage facilities, as described in claim 1, is characterized in that... The disturbance composite correction module corrects the basic feeding coordination command in two steps. The first step reads the disturbance condition classification identifier and disturbance intensity classification identifier, generates an adaptive feedforward compensation adjustment value based on the two sets of identifiers, and adds the adaptive feedforward compensation adjustment value to the basic feeding coordination command. The second step reads the furnace oxygen content value, steam pipeline pressure value, and pollutant emission concentration value output by the boiler detection subsystem, reads the steady-state operation target value corresponding to each of the three types of parameters, calculates the deviation between the real-time value and the target value of each set of parameters and the deviation change rate per unit time value, generates a fine feedback correction adjustment value based on the two sets of deviation values, adds the fine feedback correction adjustment value to the command value after the first step is completed, and the value after the addition is used as the corrected feeding control command.

6. A biomass co-firing feeding system for a coal-fired boiler utilizing existing coal storage facilities, as described in claim 1, is characterized in that... The variable load collaborative decision-making module pre-stores the safe storage threshold for raw coal as the raw coal storage mass value corresponding to 15% of the total effective volume of the coal storage silo, and pre-stores the maximum biomass storage threshold as the biomass briquette fuel storage mass value corresponding to 40% of the total effective volume of the coal storage silo. Before each generation of a coal feeder collaborative instruction, the variable load collaborative decision-making module performs a stock verification operation, reads the real-time remaining total mass value of raw coal and the real-time remaining total mass value of biomass briquettes output by the data acquisition and analysis module, and performs a size comparison operation between the two sets of stock values ​​and the two pre-stored threshold values.

7. A biomass co-firing feeding system for a coal-fired boiler utilizing existing coal storage facilities as described in claim 1, characterized in that, The variable load collaborative decision-making module classifies the variable load condition judgment value as the load change rate of 2% of the rated unit load per minute. The variable load collaborative decision-making module divides the variable load condition judgment conditions into three categories: Category 1, the load increase rate per unit time is greater than 2% of the rated unit load per minute; Category 2, the load decrease rate per unit time is greater than 2% of the rated unit load per minute, and the real-time unit load value is less than 50% of the rated unit load; Category 3, the unit load value decreases by more than 30% of the rated unit load per unit time, and the load decrease rate is less than or equal to 2% of the rated unit load per minute. The variable load collaborative decision-making module executes the increase / decrease allocation logic in the order of Category 1, Category 2, and Category 3. The variable load collaborative decision-making module internally stores the feed adjustment rate limit value as the feeder speed adjustment amount per second does not exceed 5% of the rated speed. Before generating the raw coal feeder collaborative instruction and the biomass briquette feeder collaborative instruction, a limit verification calculation is performed to verify the relationship between the unit time instruction adjustment value and the limit value.

8. A biomass co-firing feeding system for a coal-fired boiler utilizing existing coal storage facilities, as described in claim 1, is characterized in that... The closed-loop feedback optimization module receives boiler operating parameters including real-time calorific value of flue gas entering the furnace, boiler main steam pressure, boiler main steam temperature, oxygen content at furnace outlet, real-time emission concentration of flue gas pollutants, steam parameter response time under unit load change conditions, and duration of boiler combustion parameter fluctuations after feeding disturbances. The closed-loop feedback optimization module performs deviation calculations on each set of boiler operating parameters, reads the intermediate value of the target steady-state operation range corresponding to the parameter, calculates the magnitude of the deviation between the real-time parameter value and the intermediate value, the number of sampling periods in which the deviation persists, and the increase or decrease of the deviation per unit time, and stores the deviation calculation results for all sampling periods.

9. A biomass co-firing feeding system for a coal-fired boiler utilizing existing coal storage facilities, as described in claim 1, is characterized in that... After completing the threshold update, the closed-loop feedback optimization module decomposes and outputs four types of values: the first type is the stratified mixing deviation compensation coefficient, the second type is the flow fluctuation characteristic weight coefficient, the third type is the material level drop rate fluctuation weight coefficient, and the fourth type is the stratified blending ratio fluctuation coupling weight coefficient. The closed-loop feedback optimization module sets up three independent data transmission branches to complete parameter feedback. The first branch transmits the stratified mixing deviation compensation coefficient to the stratified blending ratio calculation module. The second branch transmits the flow fluctuation characteristic weight coefficient, the material level drop rate fluctuation weight coefficient, and the stratified blending ratio fluctuation coupling weight coefficient to the disturbance composite correction module. The third branch transmits the long-term cumulative boiler operating parameter deviation calculation result to the stratified blending ratio calculation module.

10. A collaborative control method for a biomass co-firing feeding system of a coal-fired boiler utilizing existing coal storage facilities, the method being applicable to the biomass co-firing feeding system of a coal-fired boiler utilizing existing coal storage facilities as described in any one of claims 1-9, characterized in that, The specific steps of this method are as follows: S1. The data acquisition and analysis module collects time-series data of material level, feed flow rate, unit load, and boiler flue gas and steam parameters at a frequency of 1Hz; it retrieves parameters of bin geometry, fuel density, and dielectric constant; and analyzes raw coal, biomass stratification interface, stacking height, and real-time inventory based on the three-dimensional material surface point cloud in the bin, and outputs stratified inventory data. S2. The layered proportioning calculation module reads the layered stock time-series data, uses the layered time-series blending ratio calculation algorithm to calculate the instantaneous blending ratio, compares it with the preset target blending ratio to obtain the deviation compensation amount, and adds it to the reference speed to generate the basic feeding coordination command. S3, the disturbance composite correction module matches three types of disturbance conditions—collapse, bridging, and low material level—through a 5-second sliding window, and calculates the comprehensive quantification value of the disturbance using a spatiotemporal coupled disturbance quantification algorithm; first, it performs feedforward compensation based on the disturbance condition, and then performs fine feedback correction based on furnace oxygen, steam pressure, and pollutant parameters, and outputs the corrected feeding control command. S4. The variable load collaborative decision-making module verifies whether the storage capacity of raw coal and biomass silos meets the storage threshold; it divides three types of variable load conditions according to the load rise and fall rate and matches the raw coal and biomass increase and decrease distribution logic; it verifies according to the limit rule that the speed adjustment per second does not exceed 5% of the rated speed and issues collaborative feeding instructions to the raw coal feeder and biomass feeder respectively. S5, the closed-loop feedback optimization module collects boiler operating parameters throughout the day and calculates the steady-state deviation of each parameter; it uses a partitioned threshold adaptive iterative algorithm to update the disturbance judgment threshold, and generates hierarchical identification adjustment parameters, mixing compensation coefficients, and disturbance weight coefficients, which are fed back to the data acquisition and analysis module, hierarchical proportioning calculation module, and disturbance composite correction module through three branches, and achieve full-process adaptive optimization through iterative loop.

Images