Intelligent Operation and Maintenance Methods, Systems, and Equipment for Coal Mills Based on RCM

Abstract

This invention belongs to the field of coal mill operation and maintenance technology, and discloses a method, system, and equipment for intelligent operation and maintenance of coal mills based on RCM (Real-Time Management). The method includes aligning collected online monitoring data and offline detection data along the time dimension to construct a multi-dimensional fusion feature set; dynamically adjusting the dimensional scores of a four-dimensional scoring model based on the equipment structure tree and fault mode library, and updating the real-time risk level of each fault mode; inputting the multi-dimensional fusion feature set into a wear prediction model to output the real-time wear amount and remaining service life of the grinding rollers; dynamically solving for the target operating parameter combination based on the real-time risk level using an operation optimization model; calculating the maintenance urgency based on the real-time risk level and remaining service life, and generating maintenance work orders and adjustment instructions; and generating standardized operation and maintenance cases based on the execution results and storing them in a case library for optimizing subsequent decision-making logic and model parameters. This approach achieves dynamic, precise, and intelligent operation and maintenance of coal mills.

Description

Technical Field

[0001] This invention relates to the field of coal mill operation and maintenance technology, and in particular to an intelligent operation and maintenance method, system and equipment for coal mills based on RCM. Background Technology

[0002] The coal mill is a core piece of equipment in the pulverizing system of a coal-fired power unit, and its operating status directly affects the unit's safety and economy. Currently, coal mill maintenance mainly relies on periodic inspections, which are often haphazard and prone to over- or under-maintenance. Existing Reliability-Centered Maintenance (RCM) applications are largely limited to static failure mode analysis, disconnected from real-time operational data, and unable to achieve dynamic risk assessment. Grinding roller wear relies on manual offline detection, lacking real-time predictive capabilities. Online operational data and offline maintenance data are independent, making integrated analysis impossible. The simplistic operating strategies are ill-suited to varying coal quality conditions, leading to high energy consumption and rapid component wear. These issues hinder the improvement of intelligent coal mill maintenance.

[0003] The above content is only used to help understand the technical solution of the present invention and does not represent an admission that the above content is prior art. Summary of the Invention

[0004] The main objective of this invention is to provide an intelligent operation and maintenance method, system, and equipment for coal mills based on RCM, aiming to solve the technical problems in existing coal mill operation and maintenance, such as static maintenance strategies, lagging risk analysis, lack of wear prediction, data silos, and insufficient optimization of operating parameters.

[0005] To achieve the above objectives, the present invention provides an intelligent operation and maintenance method for coal mills based on RCM, the method comprising the following steps: Online monitoring data of the coal mill is collected in real time at a first frequency, including vibration, temperature, pressure, flow rate and coal powder fineness data. Offline detection data is collected at a second frequency lower than the first frequency, including grinding roller wear data and maintenance text data. Align the online monitoring data and the offline detection data in the time dimension, and convert the maintenance text data into quantitative labels through natural language processing to construct a multi-dimensional fusion feature set; Based on the pre-built coal mill equipment structure tree and typical failure mode library, combined with the multi-dimensional fusion feature set, the score of at least one dimension in the preset four-dimensional failure mode scoring model is dynamically adjusted, and the real-time risk level of each failure mode is updated using the adjusted four-dimensional failure mode scoring model. The multidimensional fusion feature set is input into the wear prediction model, which outputs the real-time wear amount and the predicted remaining service life of the grinding roller. When the deviation between the real-time wear amount and the predicted value exceeds a preset threshold, the model incremental learning is automatically triggered to update the model parameters of the wear prediction model. An operation optimization model is constructed with the objective function of reducing pulverization unit consumption and the constraints of wear rate and coal powder fineness. The operation optimization model is then used to dynamically solve the target operation parameter combination based on the real-time risk level. The target operation parameter combination includes air-coal ratio, loading force, and separator baffle opening. Based on the real-time risk level and the predicted remaining service life, the maintenance urgency is calculated using a preset maintenance priority function, and a maintenance work order containing maintenance content, window period and spare parts recommendations is generated. Adjustment instructions are generated based on the target operating parameter combination and sent to the corresponding actuators for adjustment. Based on the execution results of maintenance work orders and the adjustment effects of target operating parameters, standardized operation and maintenance cases are generated. These standardized operation and maintenance cases are stored in a case library for optimizing subsequent decision-making logic and model parameters.

[0006] In one embodiment, aligning the online monitoring data with the offline monitoring data in the time dimension includes: For different types of offline detection data, corresponding multi-scale sliding time windows are set respectively. Based on the multi-scale sliding time windows, the online monitoring data in each sliding time window are aggregated into a corresponding statistical feature set. The statistical feature set includes at least one of mean, maximum value, minimum value, standard deviation, rate of change and spectral energy. Real-time monitoring of abnormal characteristic events in online monitoring data, wherein the abnormal characteristic events include at least one of vibration surge events, temperature surge events, and wear-clogging precursor characteristic events; When the abnormal feature event is detected, the sliding time window is automatically expanded, and offline detection data within a preset time range before and after the occurrence of the abnormal feature event is retrieved for correlation analysis.

[0007] In one embodiment, the four-dimensional scoring model includes a severity score, an occurrence frequency score, a detectability score, and an operational optimization condition score, with the risk level R = S + O + D + C, where S is the severity score, O is the occurrence frequency score, D is the detectability score, and C is the operational optimization condition score.

[0008] In one embodiment, dynamically adjusting the scoring weight of at least one dimension in the preset four-dimensional failure mode scoring model includes: The running time of the coal mill is extracted from the multidimensional fusion feature set, and the severity score is dynamically adjusted based on the running time of the coal mill. When the remaining service life is less than 30 days, the severity score is increased by 0.5 points, and when the remaining service life is less than 15 days, the severity score is increased by 1 point. Real-time vibration data is extracted from the multi-dimensional fusion feature set, and the occurrence frequency score is dynamically adjusted based on the effective value and peak change rate of the real-time vibration data. Specifically, when the effective vibration value exceeds 80% of the warning threshold, the occurrence frequency score is increased by 0.5 points, and when the effective vibration value exceeds the warning threshold, the occurrence frequency score is increased by 1 point. Real-time temperature data is extracted from the multi-dimensional fusion feature set, and the detectability score is dynamically adjusted based on the real-time temperature data and its changing trend. When the temperature measurement point shows abnormal fluctuations but does not exceed the limit, the detectability score is automatically reduced by 0.5 points. Coal quality data is extracted from the multidimensional fusion feature set, and the operation optimization condition score is dynamically adjusted based on the ash and moisture content in the coal quality data. When the ash content exceeds the upper limit of the designed coal type, the operation optimization condition score is increased by 0.5 points.

[0009] In one embodiment, the input features of the wear prediction model include: mill output, loading force, separator speed, air-to-coal ratio, inlet primary air volume, inlet primary air temperature, outlet temperature, inlet-outlet differential pressure, mill bowl differential pressure, vibration characteristic value, coal calorific value, coal ash content, coal moisture content, coal powder fineness, and cumulative operating time; the output of the wear prediction model includes: radial wear of the grinding roller, grinding roller wear rate, remaining service life of the grinding roller, and predicted surface profile of the grinding roller after wear; the wear prediction model adopts a multi-model fusion strategy, including a main prediction model and multiple auxiliary models. The main prediction model uses an LSTM network to capture time series features, and the auxiliary models use random forests to capture nonlinear relationships. The output is a weighted combination of the prediction results of each model, and the weights are dynamically adjusted according to the prediction error of each model. The incremental learning of the model includes: automatically triggering incremental learning when the deviation between the offline detected wear amount and the model prediction value exceeds 5% for three consecutive times. Incremental learning adopts a sliding window mechanism, using data from the most recent 6 months to fine-tune the model and retain the long-term characteristics of historical data. During incremental learning, an elastic weight consolidation algorithm is used to maintain the memory of important historical features while updating new data features. After the incremental learning is completed, the effect is automatically evaluated on the validation set. If the model performance improves, the original model is replaced; otherwise, it is rolled back to the previous version.

[0010] In one embodiment, the objective function is: min pulverizing unit consumption = f(air-coal ratio, loading force, separator baffle opening, coal quality parameters, grinding roller wear status); the constraints include at least: grinding roller wear rate ≤ design allowable maximum value × (1 + risk level), coal powder fineness within the target value ±5%, pulverizer outlet temperature within the safe operating range; pulverizer current does not exceed rated current, and grinding bowl differential pressure is within the design range; The optimization cycle of operating parameters is dynamically adjusted according to the magnitude of coal quality changes. When the coal quality fluctuation exceeds the preset threshold, the optimization cycle is shortened to 5-10 minutes; when the coal quality is stable, the optimization cycle is extended to 30-60 minutes.

[0011] In one embodiment, the maintenance priority function is: P = w1 × R_level + w2 × (1 - L_remaining / L_baseline), where P is the maintenance urgency, R_level is the real-time risk level, L_remaining is the predicted remaining service life, L_baseline is the baseline service life, and w1 and w2 are weighting coefficients. The maintenance work order includes: a list of required tools, personnel qualification requirements, safety measures reminders, and standard operating procedure guidelines. The maintenance work order is divided into four levels according to the P value: emergency maintenance (P≥8), priority maintenance (6≤P<8), planned maintenance (4≤P<6), and routine maintenance (P<4).

[0012] Furthermore, to achieve the above objectives, this invention also proposes an RCM-based intelligent operation and maintenance system for coal mills. This RCM-based intelligent operation and maintenance system is applied to the RCM-based intelligent operation and maintenance method for coal mills described above. The system includes: The data acquisition module is used to acquire online monitoring data of the coal mill in real time at a first frequency. The online monitoring data includes vibration, temperature, pressure, flow rate and coal powder fineness data. The module also acquires offline detection data at a second frequency lower than the first frequency. The offline detection data includes grinding roller wear data and maintenance text data. The data alignment and fusion module is used to align the online monitoring data and the offline detection data in the time dimension, and convert the maintenance text data into quantitative labels through natural language processing to construct a multi-dimensional fusion feature set; The dynamic risk analysis module is used to dynamically adjust the score of at least one dimension in the preset four-dimensional fault failure mode scoring model based on the pre-built coal mill equipment structure tree and typical fault mode library combined with the multi-dimensional fusion feature set, and to update the real-time risk level of each fault mode using the adjusted four-dimensional fault failure mode scoring model. The wear prediction and self-calibration module is used to input the multi-dimensional fusion feature set into the wear prediction model, output the real-time wear amount and remaining service life prediction value of the grinding roller, and automatically trigger incremental learning of the model when the deviation between the real-time wear amount and the prediction value exceeds a preset threshold, so as to update the model parameters of the wear prediction model. The operation optimization module is used to construct an operation optimization model with the objective function of reducing pulverization unit consumption and the constraints of wear rate and coal powder fineness. The operation optimization model is used to dynamically solve the target operation parameter combination based on the real-time risk level. The target operation parameter combination includes air-coal ratio, loading force and separator baffle opening. The intelligent decision-making module is used to calculate the urgency of maintenance based on the real-time risk level and the predicted value of remaining service life through a preset maintenance priority function, and generate a maintenance work order that includes maintenance content, window period and spare parts suggestions, as well as generate adjustment instructions based on the target operating parameter combination and issue them to the corresponding actuators for adjustment; The feedback optimization module is used to generate standardized operation and maintenance cases based on the execution results of maintenance work orders and the adjustment effect of target operating parameters. These standardized operation and maintenance cases are stored in a case library for optimizing subsequent decision-making logic and model parameters.

[0013] Furthermore, to achieve the above objectives, the present invention also proposes an RCM-based intelligent operation and maintenance device for coal mills. The RCM-based intelligent operation and maintenance device for coal mills includes: a memory, a processor, and an RCM-based intelligent operation and maintenance program for coal mills stored in the memory and executable on the processor. The RCM-based intelligent operation and maintenance program for coal mills is configured to implement the steps of the RCM-based intelligent operation and maintenance method for coal mills as described above.

[0014] Furthermore, to achieve the above objectives, the present invention also proposes a storage medium storing an RCM-based intelligent operation and maintenance program for a coal mill. When the RCM-based intelligent operation and maintenance program is executed by a processor, it implements the steps of the RCM-based intelligent operation and maintenance method for a coal mill as described above.

[0015] This invention aligns collected online monitoring data with offline detection data along the time dimension to construct a multi-dimensional fusion feature set. Based on the equipment structure tree and fault mode library, combined with the multi-dimensional fusion feature set, the dimensional scores of the four-dimensional scoring model are dynamically adjusted to update the real-time risk level of each fault mode. The multi-dimensional fusion feature set is input into the wear prediction model, outputting the real-time wear amount and remaining service life of the grinding rollers. The operation optimization model dynamically solves for the target operating parameter combination based on the real-time risk level. The maintenance urgency is calculated based on the real-time risk level and remaining service life, generating maintenance work orders and adjustment instructions. Standardized operation and maintenance cases are generated based on the execution results and stored in the case library for optimizing subsequent decision-making logic and model parameters. This approach achieves dynamic, precise, and intelligent operation and maintenance of the coal mill. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating the first embodiment of the intelligent operation and maintenance method for coal mills based on RCM of the present invention. Figure 2 This is a structural block diagram of the first embodiment of the intelligent operation and maintenance system for coal mills based on RCM of the present invention.

[0017] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0018] It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention.

[0019] This invention provides an intelligent operation and maintenance method for coal mills based on RCM, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of an intelligent operation and maintenance method for coal mills based on RCM according to the present invention.

[0020] In this embodiment, the RCM-based intelligent operation and maintenance method for coal mills includes the following steps: Step S10: Collect online monitoring data of the coal mill in real time at a first frequency. The online monitoring data includes vibration, temperature, pressure, flow rate and coal powder fineness data. Collect offline detection data at a second frequency lower than the first frequency.

[0021] In this embodiment, the execution subject is an RCM-based intelligent operation and maintenance device for coal mills. This RCM-based intelligent operation and maintenance device for coal mills has functions such as data processing, data communication, and program execution. The RCM-based intelligent operation and maintenance device for coal mills can be a computer terminal device or other network device, or other devices with similar functions. This embodiment does not limit this.

[0022] It should be noted that the coal mill is the core equipment of the pulverizing system in a coal-fired power unit, and its operating status directly affects the safety and economy of the unit. Currently, coal mill maintenance mainly adopts a periodic inspection approach, which is highly arbitrary and prone to over- or under-maintenance. Existing RCM (Reliability-Centered Maintenance) applications are mostly limited to static failure mode analysis, disconnected from real-time operational data, and unable to achieve dynamic risk assessment. Grinding roller wear relies on manual offline detection, lacking real-time predictive capabilities. Online operational data and offline maintenance data are independent, making integrated analysis impossible. The single operating strategy is difficult to adapt to varying coal quality conditions, resulting in high energy consumption and rapid component wear. These problems hinder the improvement of the intelligent level of coal mill maintenance.

[0023] To address the aforementioned technical challenges, this embodiment aligns the collected online monitoring data with offline detection data along the time dimension to construct a multi-dimensional fusion feature set. Based on the equipment structure tree and fault mode library, combined with the multi-dimensional fusion feature set, the dimensional scores of the four-dimensional scoring model are dynamically adjusted, updating the real-time risk level of each fault mode. The multi-dimensional fusion feature set is input into the wear prediction model, outputting the real-time wear amount and remaining service life of the grinding rollers. The operation optimization model dynamically solves for the target operating parameter combination based on the real-time risk level. The maintenance urgency is calculated based on the real-time risk level and remaining service life, generating maintenance work orders and adjustment instructions. Standardized operation and maintenance cases are generated based on the execution results and stored in the case library for optimizing subsequent decision-making logic and model parameters. This approach achieves dynamic, precise, and intelligent operation and maintenance of the coal mill. Specifically, it can be implemented as follows.

[0024] In specific implementation, the offline detection data includes grinding roller wear data and maintenance text data. The first frequency is used to collect online operating data of the coal mill in real time. It is usually set to the millisecond to second level according to the data change rate and analysis requirements. Typical value ranges are as follows: vibration data is 2560Hz-12800Hz, temperature data is 0.5Hz-1Hz (once every 1-2 seconds), pressure data is 1Hz-5Hz, flow data is 1Hz-5Hz, coal powder fineness data is 0.0167Hz-0.1Hz (once every 10-60 seconds), and environmental data is 0.0167Hz-0.033Hz (once every 30-60 seconds). The second frequency is used to collect offline data such as grinding roller wear and maintenance records. It is usually set to daily, weekly, monthly or event-triggered. Typical value ranges are as follows: grinding roller wear data once a week / once a month, maintenance text data triggered by events, coal quality data (laboratory analysis) once a day / once a week, and offline oil detection once a month / once a quarter.

[0025] Step S20: Align the online monitoring data and the offline detection data in the time dimension, and convert the maintenance text data into quantitative labels through natural language processing to construct a multi-dimensional fusion feature set.

[0026] In specific implementation, the alignment in the time dimension is achieved by setting corresponding multi-scale sliding time windows for different types of offline detection data. Based on these multi-scale sliding time windows, the online monitoring data within each sliding time window is aggregated into a corresponding statistical feature set. The statistical feature set includes at least one of the following: mean, maximum, minimum, standard deviation, rate of change, and spectral energy. Abnormal feature events in the online monitoring data are monitored in real time. These abnormal feature events include at least one of the following: sudden increase in vibration, sudden rise in temperature, and pre-clogging wear feature events. When an abnormal feature event is identified, the sliding time window is automatically expanded, and offline detection data within a preset time range before and after the occurrence of the abnormal feature event is retrieved for correlation analysis.

[0027] It should be noted that, taking the wear data of the grinding roller as an example, data is collected weekly, with each test yielding a specific wear value (unit: mm). To align the high-frequency online monitoring data with this offline detection data, the sliding time window is set as follows: Window length: 168 hours (i.e., 7 days, covering the complete cycle between two tests); Window sliding step: 24 hours (sliding once per day); Alignment method: The end time of each window (i.e., the offline detection time) is used as the alignment point, and the online monitoring data within that window is aggregated into a statistical feature set. Furthermore, taking maintenance text data as an example, this maintenance text data is event-triggered data, entered in real time after each maintenance or inspection. For this type of data, the adaptive time window is set as follows: Window length: dynamically adjusted according to the maintenance type, with a window of 720 hours (30 days) for Class A maintenance, 360 hours (15 days) for Class C maintenance, and 72 hours (3 days) for routine maintenance; Window direction: taking half the window length forward and half the window length backward from the maintenance occurrence time as the center; Alignment method: using the maintenance occurrence time as the alignment point, aggregating the online monitoring data features within the window.

[0028] Furthermore, fault mode labels are extracted from the maintenance texts. A BERT-based text classification model is used to classify each maintenance record into the corresponding fault mode. The quantified labels are as follows: For example, input: "Severe wear of grinding roller, increased temperature of grinding roller bearing, and excessive vibration value", model output: FM-01 (grinding roller wear): confidence level 0.92, FM-02 (grinding roller bearing damage): confidence level 0.78.

[0029] Step S30: Based on the pre-built coal mill equipment structure tree and typical fault mode library, combined with the multi-dimensional fusion feature set, dynamically adjust the score of at least one dimension in the preset four-dimensional fault failure mode scoring model, and use the adjusted four-dimensional fault failure mode scoring model to update the real-time risk level of each preset fault mode.

[0030] In this embodiment, the four-dimensional scoring model includes severity score, frequency of occurrence score, detectability score, and operational optimization condition score. The risk level R = S + O + D + C, where S is the severity score, O is the frequency of occurrence score, D is the detectability score, and C is the operational optimization condition score. Pre-defined fault modes, such as those in the coal mill body system, mainly include grinding roller wear, grinding roller bearing damage, grinding disc liner wear, grinding roller seal damage, grinding roller tie rod breakage, and loading frame deformation. These faults mainly manifest as decreased output, abnormal vibration, increased temperature, and coarser coal powder fineness, and in severe cases, may lead to grinding roller jamming or equipment damage. Loading system faults include hydraulic loading system leakage, hydraulic pump failure, accumulator failure, and loading force sensor failure, mainly affecting the stability and control accuracy of the loading force, leading to output fluctuations and unstable operation. Drive and transmission system faults cover reducer gear wear, reducer bearing damage, motor bearing failure, and coupling damage. Typical symptoms include sudden increase in vibration, abnormal noise, and increased oil temperature, posing a risk of transmission failure. Sealing system failures include pulverized coal leakage from the pulverizer body, blockage in the sealing air system, shaft seal wear, and sealing fan malfunctions. These primarily lead to on-site dust pollution, seal failure, and accelerated bearing wear due to pulverized coal entering the bearings. Separator system failures include separator baffle jamming, separator bearing damage, separator motor malfunctions, and internal separator wear, directly affecting the accuracy of pulverized coal fineness adjustment and separation efficiency. Pulverized coal air system failures include mill blockage, abnormal primary air volume, malfunctions in the stone and coal discharge system, and pulverized coal pipeline blockages. These mainly manifest as abnormal inlet and outlet differential pressure, output fluctuations, and decreased pulverizing efficiency. The real-time risk level of each failure mode can be calculated using the formulas described above.In one embodiment, dynamically adjusting the scoring weight of at least one dimension in the preset four-dimensional fault failure mode scoring model includes: extracting the operating time of the coal mill from the multi-dimensional fusion feature set, and dynamically adjusting the severity score based on the operating time of the coal mill, wherein when the remaining service life is less than 30 days, the severity score is increased by 0.5 points, and when the remaining service life is less than 15 days, the severity score is increased by 1 point; extracting real-time vibration data from the multi-dimensional fusion feature set, and dynamically adjusting the occurrence frequency score based on the effective value and peak value change rate of the real-time vibration data, wherein when the effective value of vibration exceeds the warning threshold... When the occurrence rate reaches 80%, the frequency score is increased by 0.5 points; when the effective vibration value exceeds the warning threshold, the frequency score is increased by 1 point. Real-time temperature data is extracted from the multi-dimensional fusion feature set, and the detectability score is dynamically adjusted based on the real-time temperature data and its changing trend. Specifically, when abnormal fluctuations occur at the temperature measuring point but do not exceed the limit, the detectability score is automatically decreased by 0.5 points. Coal quality data is extracted from the multi-dimensional fusion feature set, and the operation optimization condition score is dynamically adjusted based on the ash and moisture content in the coal quality data. Specifically, when the ash content exceeds the upper limit of the designed coal type, the operation optimization condition score is increased by 0.5 points. After completing the score adjustment, the risk level can be calculated by substituting the values ​​into the above formula. Since the score is dynamically adjusted in real time, the corresponding risk level is also updated in real time.

[0031] Step S40: Input the multidimensional fusion feature set into the wear prediction model, output the real-time wear amount and remaining service life prediction value of the grinding roller, and automatically trigger incremental learning of the model when the deviation between the real-time wear amount and the prediction value exceeds a preset threshold to update the model parameters of the wear prediction model.

[0032] It should be noted that the input features of the wear prediction model include: mill output, loading force, separator speed, air-to-coal ratio, inlet primary air volume, inlet primary air temperature, outlet temperature, inlet-outlet differential pressure, mill bowl differential pressure, vibration characteristic value, coal calorific value, coal ash content, coal moisture content, coal powder fineness, and cumulative operating time. The output of the wear prediction model includes: radial wear of the grinding roller, grinding roller wear rate, remaining service life of the grinding roller, and predicted surface profile of the grinding roller after wear. The wear prediction model adopts a multi-model fusion strategy, including one main prediction model and multiple auxiliary models. The main prediction model uses an LSTM network to capture time series features, and the auxiliary models use random forests to capture nonlinear relationships. The output is a weighted combination of the prediction results of each model, and the weights are dynamically adjusted according to the prediction error of each model.

[0033] In the specific implementation, to predict the remaining service life, this embodiment also adopts an incremental learning approach for the face model. Specifically, when the deviation between the offline detected wear amount and the model prediction exceeds 5% for three consecutive times, incremental learning is automatically triggered. The incremental learning adopts a sliding window mechanism, using data from the most recent 6 months to fine-tune the model and retain the long-term characteristics of historical data. During the incremental learning process, an elastic weight consolidation algorithm is used to maintain the memory of important historical features while updating new data features. After the incremental learning is completed, the effect is automatically evaluated on the validation set. If the model performance improves, the original model is replaced; otherwise, it is rolled back to the previous version.

[0034] Step S50: Construct an operation optimization model with the objective function of reducing pulverization unit consumption and the constraints of wear rate and coal powder fineness, and use the operation optimization model to dynamically solve the target operation parameter combination based on the real-time risk level.

[0035] In the specific implementation, the objective function is: min pulverizing unit consumption = f(air-coal ratio, loading force, separator baffle opening, coal quality parameters, grinding roller wear status); the constraints include at least: grinding roller wear rate ≤ design maximum allowable value × (1 + risk level), coal powder fineness within the target value ±5%, pulverizer outlet temperature within the safe operating range; pulverizer current not exceeding the rated current, and grinding bowl differential pressure within the design range. Based on these constraints, this embodiment uses a multi-objective particle swarm optimization algorithm to solve the operational optimization model. The optimization solution process includes: 1: Initializing the particle swarm and randomly generating 50 sets of decision variable combinations (air-coal ratio, loading force, separator baffle opening). 2: Reading the current coal quality parameters (calorific value, ash content, moisture) and grinding roller wear status (wear amount, remaining life). 3: Reading the real-time risk level output by the RCM dynamic risk analysis module and calculating the dynamic adjustment coefficient of each constraint. 4: For each particle, calculating the objective function value (pulverizing unit consumption) and the degree of constraint violation. 5: Calculating the particle fitness using the penalty function method. 6: Update the individual and global optimal values ​​for each particle. 7: Update the particle velocity and position. 8: Repeat steps 4-7 until the maximum number of iterations or convergence condition is reached. 9: Output the Pareto front, which is a set of non-dominated optimal solutions. Select the final target operating parameter combination from the Pareto front, using the following selection strategy: When the real-time risk level is low (R_risk < 0.3), prioritize the solution with the lowest pulverizing unit consumption; when the real-time risk level is low to medium (0.3 ≤ R_risk < 0.5), select the solution that balances pulverizing unit consumption and wear rate; when the real-time risk level is medium to high (0.5 ≤ R_risk < 0.7), prioritize the solution with the lowest wear rate; when the real-time risk level is high (R_risk ≥ 0.7), prioritize the conservative solution that satisfies safety constraints and trigger a maintenance warning.

[0036] Step S60: Calculate the maintenance urgency based on the real-time risk level and the predicted remaining service life using a preset maintenance priority function, generate a maintenance work order containing maintenance content, window period, and spare parts recommendations, and generate adjustment instructions based on the target operating parameter combination and issue them to the corresponding actuators for adjustment.

[0037] In this embodiment, the maintenance priority function is: P = w1 × R_level + w2 × (1 - L_remaining / L_baseline), where P is the maintenance urgency, R_level is the real-time risk level, L_remaining is the predicted remaining service life, L_baseline is the baseline service life, and w1 and w2 are weighting coefficients. The maintenance work order includes: a list of required tools, personnel qualification requirements, safety measures tips, and standard operating procedure guidelines. The maintenance work order is divided into four levels according to the P value: emergency maintenance (P≥8), priority maintenance (6≤P<8), planned maintenance (4≤P<6), and routine maintenance (P<4).

[0038] Step S70: Generate standardized operation and maintenance cases based on the execution results of the maintenance work orders and the adjustment effects of the target operating parameters.

[0039] It should be noted that the adjustment effect is calculated by collecting operating data for 48 hours before and after the adjustment and calculating the average change: Before adjustment, the pulverizing unit consumption was 28.5 kWh / t, and after adjustment, the pulverizing unit consumption was 26.2 kWh / t (a decrease of 8.1%); before adjustment, the grinding roller wear rate was 0.035 mm / thousand tons of coal, and after adjustment, the grinding roller wear rate was 0.028 mm / thousand tons of coal (a decrease of 20%); before adjustment, the coal powder fineness R90 was 26%, and after adjustment, the coal powder fineness R90 was 22% (qualified); before adjustment, the coal mill current was 72 A, and after adjustment, the coal mill current was 68 A (a decrease of 5.6%).

[0040] Furthermore, the structure of the generated standardized operation and maintenance case is as follows: Case Number: CASE-MPS-2024-003 Case Type: Combined Case of Grinding Roller Wear Control and Operation Optimization. I. Case Background: Equipment Information: Zhuozhou Thermal Power Unit 2A Coal Mill, MPS type medium-speed mill, cumulative operating time 15200 hours. Initial Problems: Grinding roller bearing temperature 82.3℃ (exceeding standard), grinding roller vibration 4.2mm / s (exceeding standard), coal powder fineness R90 (too coarse), pulverizing unit consumption 28.5kWh / t (too high). Coal Quality Conditions: Ash content of coal fed into the furnace 32%-38%, calorific value 3200-3600kCal / kg, deviating from the design coal type. RCM Analysis Results: The risk level of grinding roller wear (FM-01) is "high" (overall score 9 points), and the risk level of grinding roller bearing damage (FM-02) is "medium-high" (overall score 7 points). Standardized operation and maintenance cases are stored in a case library to optimize subsequent decision-making logic and model parameters.

[0041] In this embodiment, the collected online monitoring data and offline detection data are aligned in the time dimension to construct a multi-dimensional fusion feature set. Based on the equipment structure tree and fault mode library combined with the multi-dimensional fusion feature set, the dimensional scores of the four-dimensional scoring model are dynamically adjusted to update the real-time risk level of each fault mode. The multi-dimensional fusion feature set is input into the wear prediction model, which outputs the real-time wear amount and remaining service life of the grinding roller. The operation optimization model is used to dynamically solve the target operating parameter combination based on the real-time risk level. The maintenance urgency is calculated according to the real-time risk level and remaining service life, and maintenance work orders and adjustment instructions are generated. Standardized operation and maintenance cases are generated based on the execution results and stored in the case library for optimizing subsequent decision-making logic and model parameters. The above method realizes the dynamic, precise and intelligent operation and maintenance of the coal mill.

[0042] Furthermore, this embodiment of the invention also proposes a storage medium storing an RCM-based intelligent operation and maintenance program for a coal mill. When the RCM-based intelligent operation and maintenance program is executed by a processor, it implements the steps of the RCM-based intelligent operation and maintenance method for a coal mill as described above.

[0043] Reference Figure 2 , Figure 2 This is a structural block diagram of the first embodiment of the intelligent operation and maintenance system for coal mills based on RCM of the present invention.

[0044] like Figure 2 As shown in the figure, the intelligent operation and maintenance system for coal mills based on RCM proposed in this embodiment of the invention includes: The data acquisition module 10 is used to acquire online monitoring data of the coal mill in real time at a first frequency. The online monitoring data includes vibration, temperature, pressure, flow rate and coal powder fineness data. It also acquires offline detection data at a second frequency lower than the first frequency. The offline detection data includes grinding roller wear data and maintenance text data. The data alignment and fusion module 20 is used to align the online monitoring data and the offline detection data in the time dimension, and convert the maintenance text data into quantitative labels through natural language processing to construct a multi-dimensional fusion feature set; The dynamic risk analysis module 30 is used to dynamically adjust the score of at least one dimension in the preset four-dimensional fault failure mode scoring model based on the pre-built coal mill equipment structure tree and typical fault mode library combined with the multi-dimensional fusion feature set, and to update the real-time risk level of each fault mode using the adjusted four-dimensional fault failure mode scoring model. The wear prediction and self-calibration module 40 is used to input the multi-dimensional fusion feature set into the wear prediction model, output the real-time wear amount and remaining service life prediction value of the grinding roller, and automatically trigger incremental learning of the model when the deviation between the real-time wear amount and the prediction value exceeds a preset threshold, so as to update the model parameters of the wear prediction model. The operation optimization module 50 is used to construct an operation optimization model with the objective function of reducing pulverizing unit consumption and the constraints of wear rate and coal powder fineness, and to use the operation optimization model to dynamically solve the target operation parameter combination based on the real-time risk level. The target operation parameter combination includes air-coal ratio, loading force and separator baffle opening. The intelligent decision-making module 60 is used to calculate the urgency of maintenance based on the real-time risk level and the predicted value of remaining service life through a preset maintenance priority function, and generate a maintenance work order containing maintenance content, window period and spare parts suggestions, as well as generate adjustment instructions based on the target operating parameter combination and send them to the corresponding actuators for adjustment; The feedback optimization module 70 is used to generate standardized operation and maintenance cases based on the execution results of the maintenance work order and the adjustment effect of the target operating parameters. The standardized operation and maintenance cases are stored in the case library for optimizing subsequent decision-making logic and model parameters.

[0045] In this embodiment, the collected online monitoring data and offline detection data are aligned in the time dimension to construct a multi-dimensional fusion feature set. Based on the equipment structure tree and fault mode library combined with the multi-dimensional fusion feature set, the dimensional scores of the four-dimensional scoring model are dynamically adjusted to update the real-time risk level of each fault mode. The multi-dimensional fusion feature set is input into the wear prediction model, which outputs the real-time wear amount and remaining service life of the grinding roller. The operation optimization model is used to dynamically solve the target operating parameter combination based on the real-time risk level. The maintenance urgency is calculated according to the real-time risk level and remaining service life, and maintenance work orders and adjustment instructions are generated. Standardized operation and maintenance cases are generated based on the execution results and stored in the case library for optimizing subsequent decision-making logic and model parameters. The above method realizes the dynamic, precise and intelligent operation and maintenance of the coal mill.

[0046] This application embodiment also provides an RCM-based intelligent operation and maintenance device for coal mills, including a processor, a communication interface, a memory, and a communication bus. The processor, communication interface, and memory communicate with each other through the communication bus. The memory is used to store the RCM-based intelligent operation and maintenance program for coal mills. When the processor executes the program stored in the memory, it implements the above-mentioned RCM-based intelligent operation and maintenance method for coal mills.

[0047] The communication bus mentioned in the aforementioned RCM-based intelligent operation and maintenance equipment for coal mills can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc.

[0048] The communication interface is used for communication between the aforementioned RCM-based intelligent operation and maintenance equipment for coal mills and other equipment.

[0049] The memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.

[0050] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0051] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid state disk (SSD)).

[0052] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0053] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0054] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

[0055] It should be understood that the above are merely illustrative examples and do not constitute any limitation on the technical solution of the present invention. In specific applications, those skilled in the art can make settings as needed, and the present invention does not impose any restrictions on this.

[0056] It should be noted that the workflow described above is merely illustrative and does not limit the scope of protection of this invention. In practical applications, those skilled in the art can select some or all of the workflow to achieve the purpose of this embodiment according to actual needs, and no restrictions are imposed here.

[0057] In addition, for technical details not described in detail in this embodiment, please refer to the intelligent operation and maintenance method of coal mill based on RCM provided in any embodiment of the present invention, which will not be repeated here.

[0058] Furthermore, it should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0059] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0060] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as read-only memory (ROM) / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0061] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

[0062] It is understood that the system provided in the embodiments of the present invention corresponds to the method provided in the embodiments of the present invention, and the explanation, examples and beneficial effects of the relevant content can be referred to the corresponding parts of the above method.

Claims

1. A method for intelligent operation and maintenance of coal mills based on RCM, characterized in that, The RCM-based intelligent operation and maintenance method for coal mills includes: Online monitoring data of the coal mill is collected in real time at a first frequency, including vibration, temperature, pressure, flow rate and coal powder fineness data. Offline detection data is collected at a second frequency lower than the first frequency, including grinding roller wear data and maintenance text data. Align the online monitoring data and the offline detection data in the time dimension, and convert the maintenance text data into quantitative labels through natural language processing to construct a multi-dimensional fusion feature set; Based on the pre-built coal mill equipment structure tree and typical failure mode library, combined with the multi-dimensional fusion feature set, the score of at least one dimension in the preset four-dimensional failure mode scoring model is dynamically adjusted, and the real-time risk level of each failure mode is updated using the adjusted four-dimensional failure mode scoring model. The multidimensional fusion feature set is input into the wear prediction model, which outputs the real-time wear amount and the predicted remaining service life of the grinding roller. When the deviation between the real-time wear amount and the predicted value exceeds a preset threshold, the model incremental learning is automatically triggered to update the model parameters of the wear prediction model. An operation optimization model is constructed with the objective function of reducing pulverization unit consumption and the constraints of wear rate and coal powder fineness. The operation optimization model is then used to dynamically solve the target operation parameter combination based on the real-time risk level. The target operation parameter combination includes air-coal ratio, loading force, and separator baffle opening. Based on the real-time risk level and the predicted remaining service life, the maintenance urgency is calculated using a preset maintenance priority function, and a maintenance work order containing maintenance content, window period and spare parts recommendations is generated. Adjustment instructions are generated based on the target operating parameter combination and sent to the corresponding actuators for adjustment. Based on the execution results of maintenance work orders and the adjustment effects of target operating parameters, standardized operation and maintenance cases are generated. These standardized operation and maintenance cases are stored in a case library for optimizing subsequent decision-making logic and model parameters.

2. The intelligent operation and maintenance method for coal mills based on RCM as described in claim 1, characterized in that, Aligning the online monitoring data with the offline monitoring data in the time dimension includes: For different types of offline detection data, corresponding multi-scale sliding time windows are set respectively. Based on the multi-scale sliding time windows, the online monitoring data in each sliding time window are aggregated into a corresponding statistical feature set. The statistical feature set includes at least one of mean, maximum value, minimum value, standard deviation, rate of change and spectral energy. Real-time monitoring of abnormal characteristic events in online monitoring data, wherein the abnormal characteristic events include at least one of vibration surge events, temperature surge events, and wear-clogging precursor characteristic events; When the abnormal feature event is detected, the sliding time window is automatically expanded, and offline detection data within a preset time range before and after the occurrence of the abnormal feature event is retrieved for correlation analysis.

3. The intelligent operation and maintenance method for coal mills based on RCM as described in claim 1, characterized in that, The four-dimensional scoring model includes severity score, frequency of occurrence score, detectability score, and operational optimization condition score. The risk level R = S + O + D + C, where S is the severity score, O is the frequency of occurrence score, D is the detectability score, and C is the operational optimization condition score.

4. The intelligent operation and maintenance method for coal mills based on RCM as described in claim 3, characterized in that, The dynamic adjustment of the scoring weight of at least one dimension in the preset four-dimensional fault failure mode scoring model includes: The running time of the coal mill is extracted from the multidimensional fusion feature set, and the severity score is dynamically adjusted based on the running time of the coal mill. When the remaining service life is less than 30 days, the severity score is increased by 0.5 points, and when the remaining service life is less than 15 days, the severity score is increased by 1 point. Real-time vibration data is extracted from the multi-dimensional fusion feature set, and the occurrence frequency score is dynamically adjusted based on the effective value and peak change rate of the real-time vibration data. Specifically, when the effective vibration value exceeds 80% of the warning threshold, the occurrence frequency score is increased by 0.5 points, and when the effective vibration value exceeds the warning threshold, the occurrence frequency score is increased by 1 point. Real-time temperature data is extracted from the multi-dimensional fusion feature set, and the detectability score is dynamically adjusted based on the real-time temperature data and its changing trend. When the temperature measurement point shows abnormal fluctuations but does not exceed the limit, the detectability score is automatically reduced by 0.5 points. Coal quality data is extracted from the multidimensional fusion feature set, and the operation optimization condition score is dynamically adjusted based on the ash and moisture content in the coal quality data. When the ash content exceeds the upper limit of the designed coal type, the operation optimization condition score is increased by 0.5 points.

5. The intelligent operation and maintenance method for coal mills based on RCM as described in claim 1, characterized in that, The input features of the wear prediction model include: mill output, loading force, separator speed, air-to-coal ratio, inlet primary air volume, inlet primary air temperature, outlet temperature, inlet-outlet differential pressure, mill bowl differential pressure, vibration characteristic value, coal calorific value, coal ash content, coal moisture content, coal powder fineness, and cumulative operating time. The output of the wear prediction model includes: radial wear of the grinding roller, grinding roller wear rate, remaining service life of the grinding roller, and predicted surface profile of the grinding roller after wear. The wear prediction model adopts a multi-model fusion strategy, including one main prediction model and multiple auxiliary models. The main prediction model uses an LSTM network to capture time series features, and the auxiliary models use random forests to capture nonlinear relationships. The output is a weighted combination of the prediction results of each model, and the weights are dynamically adjusted according to the prediction error of each model. The incremental learning of the model includes: automatically triggering incremental learning when the deviation between the offline detected wear amount and the model prediction value exceeds 5% for three consecutive times. Incremental learning adopts a sliding window mechanism, using data from the most recent 6 months to fine-tune the model and retain the long-term characteristics of historical data. During incremental learning, an elastic weight consolidation algorithm is used to maintain the memory of important historical features while updating new data features. After the incremental learning is completed, the effect is automatically evaluated on the validation set. If the model performance improves, the original model is replaced; otherwise, it is rolled back to the previous version.

6. The intelligent operation and maintenance method for coal mills based on RCM as described in claim 1, characterized in that, The objective function is: min pulverizing unit consumption = f(air-coal ratio, loading force, separator baffle opening, coal quality parameters, grinding roller wear status); the constraints include at least: grinding roller wear rate ≤ design maximum allowable value × (1 + risk level), coal powder fineness within the target value ±5%, coal mill outlet temperature within the safe operating range; coal mill current does not exceed the rated current, and grinding bowl differential pressure is within the design range; The optimization cycle of operating parameters is dynamically adjusted according to the magnitude of coal quality changes. When the coal quality fluctuation exceeds the preset threshold, the optimization cycle is shortened to 5-10 minutes; when the coal quality is stable, the optimization cycle is extended to 30-60 minutes.

7. The intelligent operation and maintenance method for coal mills based on RCM as described in claim 1, characterized in that, The maintenance priority function is: P = w1 × R_level + w2 × (1 - L_remaining / L_baseline), where P is the urgency of maintenance, R_level is the real-time risk level, L_remaining is the predicted remaining service life, L_baseline is the baseline service life, and w1 and w2 are weighting coefficients. The maintenance work order includes: a list of required tools, personnel qualification requirements, safety measures tips, and standard operating procedure guidelines. The maintenance work order is divided into four levels according to the P value: emergency maintenance (P≥8), priority maintenance (6≤P<8), planned maintenance (4≤P<6), and routine maintenance (P<4).

8. A coal mill intelligent operation and maintenance system based on RCM, characterized in that, The RCM-based intelligent operation and maintenance system for coal mills is applied to the RCM-based intelligent operation and maintenance method for coal mills as described in any one of claims 1 to 7, and the system comprises: The data acquisition module is used to acquire online monitoring data of the coal mill in real time at a first frequency. The online monitoring data includes vibration, temperature, pressure, flow rate and coal powder fineness data. The module also acquires offline detection data at a second frequency lower than the first frequency. The offline detection data includes grinding roller wear data and maintenance text data. The data alignment and fusion module is used to align the online monitoring data and the offline detection data in the time dimension, and convert the maintenance text data into quantitative labels through natural language processing to construct a multi-dimensional fusion feature set; The dynamic risk analysis module is used to dynamically adjust the score of at least one dimension in the preset four-dimensional fault failure mode scoring model based on the pre-built coal mill equipment structure tree and typical fault mode library combined with the multi-dimensional fusion feature set, and to update the real-time risk level of each fault mode using the adjusted four-dimensional fault failure mode scoring model. The wear prediction and self-calibration module is used to input the multi-dimensional fusion feature set into the wear prediction model, output the real-time wear amount and remaining service life prediction value of the grinding roller, and automatically trigger incremental learning of the model when the deviation between the real-time wear amount and the prediction value exceeds a preset threshold, so as to update the model parameters of the wear prediction model. The operation optimization module is used to construct an operation optimization model with the objective function of reducing pulverization unit consumption and the constraints of wear rate and coal powder fineness. The operation optimization model is used to dynamically solve the target operation parameter combination based on the real-time risk level. The target operation parameter combination includes air-coal ratio, loading force and separator baffle opening. The intelligent decision-making module is used to calculate the urgency of maintenance based on the real-time risk level and the predicted value of remaining service life through a preset maintenance priority function, and generate a maintenance work order that includes maintenance content, window period and spare parts suggestions, as well as generate adjustment instructions based on the target operating parameter combination and issue them to the corresponding actuators for adjustment; The feedback optimization module is used to generate standardized operation and maintenance cases based on the execution results of maintenance work orders and the adjustment effect of target operating parameters. These standardized operation and maintenance cases are stored in a case library for optimizing subsequent decision-making logic and model parameters.

9. A coal mill intelligent operation and maintenance equipment based on RCM, characterized in that, The RCM-based intelligent operation and maintenance equipment for coal mills includes: a memory, a processor, and an RCM-based intelligent operation and maintenance program for coal mills stored in the memory and executable on the processor. The RCM-based intelligent operation and maintenance program for coal mills is configured to implement the steps of the RCM-based intelligent operation and maintenance method for coal mills as described in any one of claims 1 to 7.

10. A storage medium, characterized in that, The storage medium stores an RCM-based intelligent operation and maintenance program for coal mills. When the RCM-based intelligent operation and maintenance program is executed by the processor, it implements the steps of the RCM-based intelligent operation and maintenance method for coal mills as described in any one of claims 1 to 7.

Images