A waste incineration grate wear prediction and jam early warning method based on digital twinning

Abstract

The application discloses a kind of based on digital twinning waste incineration grate wear prediction and jam early warning method, belong to solid waste treatment and industrial automation monitoring technical field.The method includes: the displacement, velocity, contact force and acceleration data of grate mechanism are collected in real time by sensor group;Establish the three-dimensional geometric model of grate mechanism and define kinematic pair constraint relationship, material attribute and contact parameter, construct multi-body dynamics digital twinning model, real-time data are mapped to the model to realize virtual and real synchronous;The time-varying contact force and sliding speed signal output by digital twinning model are acquired, and the wear rate is dynamically calculated based on the modified Archard wear model and the remaining useful life is predicted, and when the prediction deviation exceeds the threshold, the dynamic wear coefficient is adaptively corrected;Real-time monitoring of grate motion parameters and comparison with normal working condition benchmark value, according to stroke deviation, speed fluctuation, acceleration peak value, trigger jam early warning signal gradually;The wear amount and remaining life of each grate piece are summarized, and a health assessment report and differentiated repair recommendations are generated.The application time-varying coupling of digital twinning and wear prediction is realized, the wear rate is calculated dynamically step by step, and three-level grading early warning is realized, which can improve the wear prediction accuracy of waste incineration grate and the early warning ability of jam failure.

Description

Technical Field

[0001] This invention relates to the field of solid waste treatment and industrial automation monitoring technology, specifically to a method for predicting wear and providing early warning of faults in the core moving parts of a mechanical grate incinerator in a waste-to-energy plant. Background Technology

[0002] Waste incineration is one of the main methods for the harmless treatment of municipal solid waste. Mechanical grate incinerators are widely used in waste-to-energy plants in China due to their good material adaptability and stable operation. The grate, as a key component that carries and reciprocates the waste for combustion, operates under high temperature, heavy load, and corrosive gas environments for extended periods. Its operational reliability directly affects incineration efficiency and equipment safety. Improving the operational reliability and equipment management efficiency of waste incineration equipment through digital and intelligent technologies has become an important direction for technological advancement in the industry.

[0003] Existing technologies have made valuable explorations in incinerator monitoring and early warning. For example, patent CN113921096B discloses a method for quantitatively detecting the wear life of grate bars by establishing a wear calculation model, which uses static load parameters and the total friction distance over a preset time period. Patent CN202323553818.X discloses a detection device that directly measures the wear depth of grate bars using a detection probe and sensor. Patent CN121383201A discloses a fault early warning method for incinerator systems based on the FastABOD algorithm, achieving anomaly detection through data-driven approaches. Patent CN121025466B discloses a method for judging grate jamming and automatically handling emergencies based on real-time acquisition of motion parameters. Furthermore, multibody dynamics analysis and wear modeling theory have been applied in mechanical fields such as aircraft landing gear.

[0004] There is still room for improvement in practical applications. For the moving parts of the grate in a mechanical grate incinerator, the load on the grate during actual operation fluctuates dynamically due to uneven distribution of waste material, changes in grate speed and acceleration, and the influence of high-temperature environment. The working environment of the grate bars is characterized by a large waste load in the drying section (mainly bearing mechanical wear), high temperature in the combustion section (mainly bearing high-temperature wear), and solid particles trapped in the gaps in the burnout section (mainly bearing abnormal wear such as uneven wear). However, existing wear detection methods either use static material mechanical parameters or rely on physical detection devices. Their wear models use fixed load parameters or can only detect the amount of wear that has occurred, making it difficult to reflect the influence of time-varying characteristics on the wear accumulation process. Existing fault early warning methods for incinerators based on digital twins mainly rely on the anomaly detection of sensor data. They have not yet established a simulation model for the multibody dynamics behavior of the grate moving parts, nor do they have the technical means to couple the digital twin simulation model with the wear prediction model. Existing technologies lack the ability to coordinate wear degradation prediction and motion jamming early warning, making it difficult to detect jamming faults caused by wear accumulation in their early stages.

[0005] This invention aims to provide a method for predicting wear and early warning of jamming in waste incinerator grates based on digital twins. By establishing a digital twin model of the grate mechanism and outputting time-varying dynamic parameters as dynamic inputs to the wear prediction model, the accuracy of wear prediction and the ability to provide early warning of jamming faults can be further improved. Summary of the Invention

[0006] The technical problem to be solved by this invention is: addressing the issues that existing waste incineration mechanical grate wear detection methods rely on static parameters and are difficult to reflect the influence of time-varying loads on the wear process, as well as the problems that existing digital twin fault early warning methods do not involve multibody dynamics simulation of grate moving parts and early warning of jamming faults, this invention provides a waste incineration grate wear prediction and jamming early warning method based on digital twins.

[0007] To achieve the above objectives, the present invention adopts the following technical solution:

[0008] A method for predicting wear and providing early warning of jamming in waste incinerator grate based on digital twins includes the following steps:

[0009] Step 1: Physical entity layer data acquisition. Real-time operational data is acquired through sensor arrays installed at key locations on the grate mechanism. These sensor arrays include non-contact displacement and velocity sensors installed at the grate drive shaft end, thin-film force sensors embedded between the grate frame and the fixed guide rail support surface, and acceleration sensors installed at the drive beam support position connecting the grate frame and the furnace side wall. Each sensor continuously acquires displacement, velocity, contact force, and acceleration data at a sampling frequency of 500Hz to 1000Hz.

[0010] Displacement sensors have been successfully applied in real-time position signal feedback for sliding and tilting grates in waste incinerators. For example, position signals are acquired by mounting photosensitive elements and laser emitters on the housing of the hydraulic cylinder controlling the grate movement. Existing DCS control systems can adjust the operating frequency by acquiring signals from on-site displacement sensors in the grate feeding device.

[0011] Step 2: Construct a digital twin model. Based on the grate furnace design drawings, establish a three-dimensional geometric model of the grate mechanism. On this three-dimensional geometric model, define the kinematic pair constraints, material properties, and contact parameters to establish a multibody dynamics simulation model of the grate mechanism. The kinematic pair constraints include the sliding pair constraints between the movable grate plate and the fixed frame, the rotational pair constraints between the hydraulic cylinder drive point and the movable grate, and the fixed pair constraints between each fixed component. Material properties include a density of 7800 kg / m³, an elastic modulus of 210 GPa, and a Poisson's ratio of 0.3. Contact parameters include a contact stiffness of 1e6 N / m and a normal operating condition sliding friction coefficient of 0.35. The real-time operating data collected in Step 1 is normalized and mapped to the multibody dynamics simulation model at a frequency of once to twice per second using a data mapping module, achieving virtual-real synchronization.

[0012] Step 3: Wear Dynamics Prediction. Obtain the time-varying contact force F_contact(t) and sliding velocity v_slide(t) output from the multibody dynamics simulation model in the digital twin model; set the simulation time step to 0.005 seconds to 0.02 seconds; calculate the wear rate based on the modified Archard wear model:

[0013] dQ / dt = (K_s(t) · F_contact(t) · v_slide(t)) / H(t)

[0014] Where dQ / dt is the wear rate, K_s(t) is the dynamic wear coefficient (initially calibrated to 1.5×10^{-5} to 8.5×10^{-5} based on material wear test data), and H(t) is the material hardness (the hardness of the grate material can be taken as 180 HB to 230 HB; for example, when using ZG40Cr25Ni20 heat-resistant cast steel, its as-cast Brinell hardness is approximately 180-230 HB). When the cumulative deviation between the simulated predicted wear amount and the actual detected wear amount exceeds the preset threshold (the preset threshold is 5%), K_s(t) is adaptively corrected. Based on the current cumulative wear amount Q_accumulated(t) and the design maximum allowable wear amount Q_max (which can be taken as 3.5 mm, measured by the wear depth in the wall thickness direction), the remaining service life RUL(t) is calculated.

[0015] RUL(t) = (Q_max - Q_accumulated(t)) / (dQ / dt)

[0016] Step 4: Graded Early Warning System. Real-time monitoring of the grate mechanism's motion parameters, including stroke change rate, speed fluctuation amplitude, and acceleration anomalies. First-level warning: Calculate the actual change in grate stroke Δs_real for a single stroke and compare it with the baseline stroke Δs_base determined by the multibody dynamics simulation model under fault-free conditions. A primary warning signal is triggered when the absolute value of the relative deviation exceeds a first preset threshold (range 10% to 15%). Second-level warning: When the stroke deviation condition is met, calculate the speed fluctuation index V_fluctuation = σ_v / μ_v, where σ_v is the standard deviation of the sliding speed signal and μ_v is the mean speed signal. A medium-level warning signal is triggered when V_fluctuation exceeds a second preset threshold (range 20% to 30%). Third-level warning: An emergency warning signal is triggered when the instantaneous peak value a_peak detected by the acceleration sensor exceeds a third preset threshold (range 2.0g to 3.0g). The normal operating condition baseline value is determined by the multibody dynamics simulation model under fault-free conditions.

[0017] In step 4, the wear prediction module and the wear prediction module form a data collaborative processing relationship: the wear trend data output by the wear prediction module can be used as one of the reference bases for the wear prediction module to set the dynamic warning threshold. That is, the first-level warning and second-level warning thresholds in the wear acceleration stage can be adaptively adjusted according to the change of wear accumulation rate; at the same time, the abnormal acceleration peak value monitored by the wear prediction module can provide feedback data for the historical deviation correction of the wear prediction module.

[0018] Step 5: Life Management and Maintenance Decisions. Summarize the real-time wear amount Q_accumulated(t) and the predicted remaining service life RUL(t) of each grate plate to generate an overall health status assessment report for the grate mechanism. Based on the differentiated distribution of the remaining service life of each component, generate differentiated maintenance recommendations and plans.

[0019] In step 3, the wear prediction model and the digital twin model form a data coupling relationship: the digital twin model provides the wear prediction model with time-varying contact force and sliding speed input; when the prediction deviation exceeds the threshold, the dynamic wear coefficient is adaptively corrected, and the corrected coefficient is re-inputted into the digital twin model to form a closed-loop feedback.

[0020] In step 5, the output of the wear prediction model is directly used to calculate the remaining service life, and the three-level warning signal output by the jamming warning module is used as a weight parameter for generating differentiated maintenance recommendations. The above collaborative relationships together form a complete technical solution encompassing "digital twin simulation → dynamic wear prediction → jamming graded warning → remaining service life management".

[0021] Beneficial effects

[0022] Compared with the prior art, the present invention has the following beneficial effects:

[0023] Theoretical analysis shows that after introducing the time-varying data output by the digital twin model into the grate condition monitoring method, the wear prediction model can dynamically adjust the wear rate calculation results according to the real-time contact force and sliding speed. This overcomes the technical defect of the traditional static wear model, which cannot reflect the dynamic fluctuation of the load in actual operation due to the assumption of time invariance of load, and realizes the point-by-point dynamic calculation of wear rate.

[0024] By integrating four functional steps—digital twin, wear prediction, jamming warning, and lifespan management—a complete closed loop of data processing and feedback control is formed. Theoretical analysis shows that the coupling relationship and control logic on which this closed-loop method is based are a feasible path for the four core functions to operate in a coordinated manner within the same methodological framework. The dynamic coupling of time-varying contact force and sliding velocity must rely on the output of multibody dynamics simulation; the real-time calculation of wear rate step by step requires the simulation support of digital twin; and the dynamic differentiated decision-making of lifespan management depends on the real-time status input of wear prediction and jamming warning. The four functions are interdependent and mutually supportive, and their overall technical effect exceeds the simple superposition of the individual functions.

[0025] Simulation results indicate that this invention can provide early warning of grate wear trends, and the wear rate calculation results can more accurately reflect the coupled influence of contact force temporal fluctuations and sliding speed time-varying distribution on wear. The jamming warning step, through a three-level graded judgment mechanism, can form a progressive warning from stroke anomalies to speed fluctuations and then to acceleration peaks, providing graded warning responses with different levels of urgency.

[0026] The sensor arrays used in this invention are all mature products in the existing industrial automation field. In terms of engineering deployment, the method of this invention is compatible with the existing automated control systems of waste-to-energy plants based on programmable logic controllers / distributed control systems. It can be connected to existing data acquisition and monitoring control systems by configuring data mapping interfaces, requiring only the addition of sensor deployments and edge computing nodes, without the need for large-scale modifications to existing production lines. This demonstrates good system compatibility and engineering feasibility.

[0027] Differences from existing technologies

[0028] Distinguishing Feature 1: The time-varying coupling method between the digital twin simulation model and the wear prediction model. In contrast, the wear model in Reference 1 (CN113921096B) uses static load W and the total friction distance L over a preset time period as input parameters, failing to reflect the time-varying characteristics of the load. Let the wear amount of the traditional static wear model be Q_static = K_s·W·(v_slide·Δt), where W is a fixed load parameter and Δt is the total duration of the preset time interval. This model assumes that the contact force and sliding speed are constant within the Δt time interval. However, the actual contact force F_contact(t) and sliding speed v_slide(t) are strictly non-constant within the Δt interval. The dynamic wear method of this invention discretizes Δt into m time steps, and the wear accumulation Q_accumulated is the algebraic sum of the instantaneous wear amounts at all steps. Simulation results are expected to show that when the load fluctuation exceeds the average of 18%, the relative error of the traditional static model can reach over 30%. This invention uses the time-varying contact force and sliding velocity output from the multibody dynamics simulation model in the digital twin model as the dynamic input to modify the Archard wear model, thereby realizing the dynamic calculation of the wear rate step by step over time and eliminating static model errors in principle.

[0029] Distinguishing Feature Two: A dedicated digital twin model for the moving parts of the waste incinerator grate. In contrast, Reference 2 (CN121383201A) is applicable to data-driven anomaly detection of the entire incinerator system, but does not establish a physical simulation model for the moving parts of the incinerator; Reference 3 (CN202311610309.0) is applicable to regenerative thermal incinerators, but also does not establish a multibody dynamics model for the mechanical moving parts. This invention establishes a dedicated multibody dynamics digital twin model for the moving mechanism of the waste incinerator grate. The definition of kinematic pair constraints (sliding pairs, rotating pairs, fixed pairs), material property settings, and contact parameter values ​​are all based on the actual operating characteristics of the grate. Furthermore, the specification clearly states that sliding pairs can be implemented using a slider-guide rail or roller-guide rail structure.

[0030] Distinguishing Feature 3: A graded jamming early warning mechanism based on motion parameters. While reference 4 (CN121025466B) involves grate jamming detection, it is a real-time fault assessment based on motion parameters and does not involve virtual simulation or wear prediction based on digital twins; references 2 and 3 do not design a dedicated early warning module for grate jamming. This invention comprehensively applies the dual functions of wear prediction and graded jamming early warning, providing graded early warning responses for different fault stages from initial anomalies to severe jamming.

[0031] Distinguishing Feature 4: System Integration of Wear Prediction and Jamming Early Warning. In existing technologies, wear life prediction and fault early warning belong to different technical solutions. This invention integrates wear degradation prediction and jamming fault early warning functions under the same methodological framework, forming a full-chain technical solution of "digital twin simulation → dynamic wear prediction → jamming graded early warning → remaining life management", and there is a clear feedback and collaborative relationship between steps 3-5. Attached Figure Description

[0032] Figure 1 This is a block diagram of the overall architecture of the method of the present invention (relationship between data flow and functional modules).

[0033] Figure 2 This is a flowchart illustrating the construction process of the digital twin model of this invention.

[0034] Figure 3 This is a flowchart of the dynamic calculation of wear rate and prediction of remaining life in this invention.

[0035] Figure 4 This is a flowchart of the logic for graded early warning of jamming faults in this invention.

[0036] Figure 5 This is a time-series curve of contact force-sliding speed in the simulation verification of this invention.

[0037] Figure 6 This is a graph showing the relationship between cumulative wear and running time in the simulation verification of this invention. Detailed Implementation

[0038] To make the technical means of implementing the present invention easier to understand, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only for explaining the relevant invention and are not intended to limit the invention.

[0039] Example 1: Method Execution Flow (Taking a typical waste-to-energy plant as an example)

[0040] This embodiment uses a 500t / d mechanical grate incinerator in a waste-to-energy plant as an application scenario. 500t / d is a standardized model of mechanical grate municipal solid waste incinerators, widely used in the industry—for example, a single 500t / d incineration line can increase the daily processing capacity of municipal solid waste from several hundred tons to over a thousand tons, and some projects use a 2×500t / d configuration (two 500t / d mechanical grate incinerators, with a total processing capacity of 1000t / d). The grate incinerator consists of five parallel grate units, each containing 14 movable grate bars and a corresponding fixed support structure. The grate bars are made of heat-resistant cast steel (typical material grade such as ZG40Cr25Ni20, with a cast Brinell hardness of approximately 180-230 HB) and are driven by hydraulic cylinders for reciprocating motion.

[0041] Step 1 is executed as follows: a displacement sensor collects displacement data x(t) of the grate bars, a velocity sensor collects sliding velocity v_slide(t), a force sensor collects contact force F_contact(t), and an acceleration sensor collects vibration data. Each sensor continuously collects data at a sampling frequency of 1000Hz and transmits the data to the data acquisition server via Ethernet.

[0042] Step 2 execution (digital twin model construction):

[0043] Based on the three-dimensional CAD model of the grate mechanism, the model is imported into multibody dynamics simulation software (such as RecurDyn).

[0044] Define kinematic pair constraints: sliding pairs (between the movable grate and the fixed frame), revolute pairs (between the hydraulic cylinder drive point and the movable grate), and fixed pairs (between the various fixed components). Sliding pairs can be implemented in actual structures using a slider-guide rail or roller-guide rail configuration, revolute pairs use pin connections, and fixed pairs are achieved through bolts or welding.

[0045] Material properties are set as follows: density 7800 kg / m^3, elastic modulus 210 GPa, Poisson's ratio 0.3.

[0046] Define contact parameters: normal contact stiffness 1e6 N / m, Coulomb friction model, static friction coefficient 0.4, and dynamic friction coefficient 0.35 under normal operating conditions.

[0047] Data mapping: After normalizing the real-time sensor data, it is input into the simulation model once per second to achieve virtual-real synchronization.

[0048] Step 3 execution: The multibody dynamics simulation model runs in a virtual-real synchronous state, outputting time-varying parameters F_contact(t) and v_slide(t), with a time step of 0.01 seconds and a single simulation duration of 14400 time steps (corresponding to 4 hours of actual running time). The wear rate is calculated using a modified Archard wear model: dQ / dt = (K_s(t)·F_contact(t)·v_slide(t)) / H(t). The initial value of the dynamic wear coefficient K_s(t) can be calibrated based on the wear test data of the grate material. Every 4 hours of operation, the simulated predicted wear amount is compared with the actual wear amount obtained through periodic inspections (such as thickness measurement during shutdown). If the cumulative deviation exceeds 5%, K_s(t) is compensated and corrected. The remaining service life RUL(t) is calculated based on the difference between Q_accumulated(t) and Q_max, where Q_max can be set to 3.5 mm.

[0049] Step 4 is executed as follows: Real-time monitoring of displacement, velocity, and acceleration is performed, and a three-level threshold is used to determine the following: a first-level warning is triggered when the relative deviation exceeds 12%; a second-level warning is triggered when the velocity fluctuation index exceeds 25%; and a third-level warning is triggered when the peak acceleration exceeds 2.5g.

[0050] Step 5: Summarize the current wear and remaining service life prediction of each grate, generate a health status assessment report, and list the components with shorter remaining service life that are recommended for priority maintenance.

[0051] Example 2: Simulation Verification

[0052] In this embodiment, the expected performance of the method is verified through a digital simulation platform when real-world testing is not available.

[0053] Simulation platform: A joint simulation platform of MATLAB / Simulink and RecurDyn was used. RecurDyn was used to solve the multibody dynamics of the grate mechanism, and MATLAB was used for wear prediction model calculation and data visualization.

[0054] Simulation parameter settings: The total simulation duration is set to 800 days of continuous operation (time step 0.01 seconds), divided into three operating condition stages to conform to the wear pattern of the "bathtub curve" of industrial equipment.

[0055] Simulation parameter settings: The total simulation duration is set to 800 days of continuous operation (time step 0.01 seconds), divided into three operating condition stages to conform to the wear pattern of the "bathtub curve" of industrial equipment.

[0056] Simulation parameter settings: The total simulation duration is set to 800 days of continuous operation (time step 0.01 seconds), divided into three operating condition stages to conform to the wear pattern of the "bathtub curve" of industrial equipment.

[0057] (a) Stable operation period (0 days to 600 days): The contact force F_contact follows a normal distribution N (μ=25kN, σ=3kN), and the sliding speed v_slide follows a normal distribution N (μ=0.15m / s, σ=0.02m / s). The wear rate is low and stable during this stage.

[0058] (b) Wear manifestation period (600 to 750 days): As the fit clearance increases, the average contact force gradually increases to 32kN, the sliding speed fluctuation amplitude σ_v increases to 0.045m / s, and the dynamic wear coefficient K_s(t) begins to show a slight nonlinear increase.

[0059] (c) Late stage of wear accumulation (750 to 800 days): The average contact force further increases to 38 kN, and the sliding speed begins to show periodic short-term sudden drops (simulating the precursor of local jamming, the speed drops to below 0.02 m / s instantaneously, lasting for about 2-5 seconds).

[0060] Dataset Description: The operating parameters (force, velocity, fluctuation range) used in the simulation are statistically scaled down based on historical operating data of waste incineration power plants with similar processing capacity to ensure that the load spectrum and velocity spectrum are consistent with the actual engineering in terms of magnitude and fluctuation characteristics.

[0061] Theoretical Derivation and Error Convergence Analysis: Assume the actual cumulative wear amount Q_true(t) satisfies the differential equation dQ_true / dt = K_s(t)·F_contact(t)·v_slide(t) / H(t), where K_s(t) is unknown. This method uses a dynamic wear model to calculate the wear amount Q_model(t), establishing the error e(t) = Q_true(t) - Q_model(t). The adaptive correction rule adjusts K_s(t) when |e(t) / Q_model(t)|>5%. Theoretical analysis shows that this adaptive correction mechanism allows the error to converge to a stable value with increasing iterations, and the final simulation convergence results in a controllable error within a small range.

[0062] Expected simulation results:

[0063] like Figure 5 As shown, during the stable operation period, the contact force is stable at 22-28kN and the speed is 0.12-0.18m / s; during the wear manifestation period, the contact force rises to 28-36kN and the speed fluctuation range increases; in the later stage of wear accumulation, the contact force remains high (36-42kN) and a significant periodic drop appears in the speed curve.

[0064] like Figure 6 As shown, the wear accumulation curve exhibits an upward convex growth trend. The wear rate is low in the initial stage (0-600 days), gradually increases in the middle stage (600-750 days), and further increases in the later stage (750-800 days), with the remaining service life decreasing rapidly. When the wear accumulation reaches approximately the maximum allowable value (e.g., 3.5 mm), it is recommended to arrange for grate replacement and maintenance.

[0065] Before the model enters the late stage of wear accumulation, the wear rate curve begins to show a recognizable nonlinear upward trend, but the wear accumulation amount does not immediately reach the threshold; the jamming warning module triggers primary, intermediate and emergency warnings in sequence according to the gradual over-limit of displacement deviation, velocity fluctuation and acceleration peak; the wear accumulation prediction error can converge to within 5% during the wear manifestation period.

[0066] Simulation results indicate that the method of this invention can accurately capture the dynamic characteristics of wear acceleration and jamming precursors in the contact force-velocity time-series signal. The dynamic prediction results of wear accumulation are consistent with the wear data trend calibrated by material tests, which can provide an effective quantitative basis for grate life management and maintenance decisions.

Claims

1. A method for predicting wear and predicting jamming in waste incinerator grates based on digital twins, characterized in that, Includes the following steps: Step A: Real-time data collection of displacement, velocity, contact force, and acceleration is achieved through sensor arrays installed at key parts of the grate mechanism; Step B: Based on the grate furnace design drawings, establish a three-dimensional geometric model of the grate mechanism, define the kinematic pair constraint relationship, material properties and contact parameters to establish a multibody dynamics simulation model, and map the real-time data collected in Step A to this simulation model to achieve virtual-real synchronization and build a digital twin model; Step C: Obtain the time-varying contact force and sliding velocity signals output by the multibody dynamics simulation model in the digital twin model, dynamically calculate the wear rate based on the modified wear model, and predict the remaining service life; Step D: Monitor the motion parameters of the grate mechanism in real time. When the deviation of the motion parameters from the normal operating condition reference value exceeds the preset threshold, generate a graded jamming warning signal. Step E: Summarize the real-time wear and remaining service life predictions of each grate, and generate a health status assessment report and differentiated maintenance recommendations.

2. The method according to claim 1, characterized in that, The sensor group described in step A includes: a non-contact displacement sensor and a velocity sensor installed at the end of the grate drive shaft, a thin-film force sensor embedded between the grate frame and the fixed guide rail support surface, and an acceleration sensor installed at the support position of the drive beam connecting the grate frame and the side wall of the furnace body.

3. The method according to claim 1, characterized in that, The kinematic pair constraints mentioned in step B include the sliding pair constraints between the movable grate and the fixed frame, the rotational pair constraints between the hydraulic cylinder drive point and the movable grate, and the fixed pair constraints between each fixed component.

4. The method according to claim 1, characterized in that, The modified wear model described in step C uses dynamic wear coefficient, time-varying contact force and sliding speed, and grate material hardness as input parameters. The dynamic wear coefficient is adaptively corrected based on the cumulative deviation between the simulated predicted wear amount and the actual detected wear amount.

5. The method according to claim 1, characterized in that, The motion parameters mentioned in step D include at least one of the following: stroke change rate, speed fluctuation amplitude, and acceleration anomaly value; the graded warning signals include: a primary warning signal triggered when the relative deviation between the actual stroke change and the normal working condition reference stroke exceeds a first preset threshold, an intermediate warning signal triggered when the speed fluctuation index exceeds a second preset threshold, and an emergency warning signal triggered when the instantaneous peak value of acceleration exceeds a third preset threshold.

6. The method according to claim 1, characterized in that, The differentiated maintenance recommendations mentioned in step E are generated based on the ranking and degree of difference of the remaining service life of each component, and priority is given to arranging maintenance for components with shorter remaining service life.

Images