Equipment health assessment method and device based on AI and storage medium

Abstract

The invention relates to the technical field of slag salvaging equipment data processing, in particular to an AI-based equipment health assessment method and device and a storage medium, and the method comprises the steps: collecting guide wheel data, chain data and slag water data of slag salvaging equipment in real time; analyzing a guide wheel torque fluctuation coefficient, a chain pitch variation rate and a slag-water dynamic balance index according to guide wheel data, chain data and slag-water data of the slag salvaging equipment; performing feature analysis on the guide wheel torque fluctuation coefficient, the chain pitch variation rate and the slag-water dynamic balance index to obtain health feature parameters; constructing a health assessment model based on the health feature parameters to analyze an equipment health index; making an evaluation decision tree based on the equipment health index; and refining the assessment decision tree according to the chain pitch variation rate, the health characteristic parameters and the slag-water dynamic balance index. According to the invention, accurate evaluation of the health state of the slag salvaging equipment is realized.

Description

Technical Field

[0001] This invention relates to the field of data processing technology for slag removal equipment, and in particular to an AI-based method, apparatus, and storage medium for equipment health assessment. Background Technology

[0002] As a key piece of equipment in industrial water treatment and power generation, the health status of slag removal equipment directly affects the system's operating efficiency and safety. Traditional manual inspections rely on experience and are prone to missed inspections. With the development of technologies such as the Internet of Things and machine learning, intelligent evaluation systems based on multi-dimensional data fusion are becoming the trend, enabling fault early warning and precise maintenance.

[0003] Current health assessments of slag removal equipment mainly rely on regular manual inspections and single-parameter threshold alarms, which have the following drawbacks: Assessment lag: Long manual inspection cycles make it impossible to capture transient anomalies in real time, resulting in delayed fault warnings; High false alarm rate: Single-parameter threshold methods ignore the coupling relationship of multiple parameters and are prone to false alarms due to environmental interference; One-sided decision-making: Traditional methods only output binary results of normal and abnormal, lacking the ability to classify health status and locate root causes, and cannot guide accurate maintenance. Summary of the Invention

[0004] The purpose of this invention is to provide an AI-based device health assessment method, apparatus, and storage medium to solve at least one of the problems existing in the prior art.

[0005] To achieve the above objectives, the present invention adopts the following technical solution:

[0006] An AI-based device health assessment method includes:

[0007] Real-time data collection of guide wheel, chain, and slag / water data from the slag removal equipment;

[0008] Based on the guide wheel data, chain data, and slag-water data of the slag removal equipment, the guide wheel torque fluctuation coefficient, chain pitch variation rate, and slag-water dynamic balance index were analyzed.

[0009] Characteristic analysis was performed on the guide wheel torque fluctuation coefficient, chain pitch variation rate, and slag-water dynamic balance index to obtain health characteristic parameters;

[0010] A health assessment model is constructed based on health characteristic parameters to analyze the equipment health index;

[0011] Develop an assessment decision tree based on equipment health index;

[0012] The evaluation decision tree is refined based on the chain pitch variation rate, health characteristic parameters, and sludge-water dynamic balance index.

[0013] Furthermore, the average torque of the guide wheel within 10 minutes is taken as the torque mean, and the guide wheel torque fluctuation coefficient is analyzed based on the guide wheel torque, torque mean, number of guide wheel turns, and number of sealant grease life turns; the chain pitch variation rate is analyzed based on the chain scraper distance and design pitch; and the slag-water dynamic balance index is analyzed based on the liquid level height, slurry density, flow rate, and slag feed rate.

[0014] Furthermore, guide wheel data satisfying TF(i)≥tf are extracted as sample guide wheel data, where tf represents the fluctuation coefficient extraction threshold, TF(i) represents the guide wheel torque fluctuation coefficient, and i represents the time number. Regression analysis is performed on the sample guide wheel data to obtain the first regression parameter and the second regression parameter. The sealing performance index is then analyzed based on the guide wheel torque fluctuation coefficient, the first regression parameter, and the second regression parameter. The expression for the sealing performance index is: SE(i)=1 / [1+e -a×(TF(i)-b) ], where SE(i) represents the sealing performance index.

[0015] Furthermore, the average value of the chain load current within 10 minutes is used as the average current parameter, and the wear acceleration is analyzed based on the average current parameter and the chain pitch variation rate. The expression for the wear acceleration is: WA(i)=d(PV(i)) / d(i)×Ravg, where WA(i) represents the wear acceleration, Ravg represents the average current parameter, d(PV(i)) / d(i) represents the derivative of the chain pitch variation rate at time number i, and PV(i) represents the chain pitch variation rate.

[0016] Furthermore, the risk deviation is analyzed based on the slag-water dynamic balance index. The expression for the risk deviation is: RD(i)=|DB(i)-1|×d(DB(i)) / d(i), where RD(i) represents the risk deviation, d(DB(i)) / d(i) represents the derivative of the slag-water dynamic balance index at time number i, and DB(i) represents the slag-water dynamic balance index.

[0017] Furthermore, a health feature vector is constructed based on health feature parameters, and the health covariance matrix is ​​analyzed based on the health feature vectors stored over 30 days. The device health index of the health covariance matrix and health feature vectors is analyzed using the Mahalanobis distance algorithm.

[0018] Furthermore, an evaluation decision tree is developed based on the equipment health index. If the equipment health index is less than the first health threshold, the equipment is determined to be in a healthy state; if the equipment health index is greater than or equal to the first health threshold and less than the second health threshold, the equipment is determined to be in a sub-healthy state; if the equipment health index is greater than or equal to the second health threshold and less than the third health threshold, the equipment is determined to be in a high-risk state; and if the equipment health index is greater than or equal to the third health threshold, the equipment is determined to be in a critical fault state.

[0019] Furthermore, the evaluation decision tree is refined based on the chain pitch variation rate, health characteristic parameters, equipment health index, and slag-water dynamic balance index. When the chain pitch variation rate is greater than the variation rate threshold and the wear acceleration is greater than the wear threshold, the equipment health status is determined to be chain abnormal. If the ratio of the standard deviation to the average value of the slag-water dynamic balance index within 10 minutes is greater than the fluctuation threshold, the equipment health status is determined to be overflow abnormal.

[0020] On the other hand, the present invention also provides an AI-based device health assessment apparatus, comprising:

[0021] The data acquisition unit is used to collect real-time data from the guide wheel, chain, and slag / water of the slag removal equipment.

[0022] The parameter analysis unit is used to analyze the guide wheel torque fluctuation coefficient, chain pitch variation rate, and slag-water dynamic balance index based on the guide wheel data, chain data, and slag-water data of the slag removal equipment.

[0023] The feature analysis unit is used to perform feature analysis on the guide wheel torque fluctuation coefficient, chain pitch variation rate and slag-water dynamic balance index to obtain health feature parameters;

[0024] The model building unit is used to build a health assessment model based on health characteristic parameters in order to analyze the equipment health index;

[0025] Decision-making unit, used to develop an assessment decision tree based on equipment health index;

[0026] The decision refinement unit is used to refine the evaluation decision tree based on the chain pitch variation rate, health characteristic parameters, and slag-water dynamic balance index.

[0027] On the other hand, the present invention also provides a storage medium characterized in that it stores instructions that, when run on a computer, cause the computer to perform the AI-based device health assessment method as described in any of the preceding claims.

[0028] The beneficial effects of this invention are as follows: By integrating multi-source data and dynamic index modeling, it achieves comprehensive real-time monitoring of mechanical transmission, sealing systems, and slag-water balance. It utilizes health characteristic parameters to transform physical signals into quantifiable indicators of equipment health, supporting early warning of progressive deterioration. Combined with multi-level evaluation decision trees and root cause refinement rules, it outputs hierarchical conclusions from overall health status to specific fault types, improving the targeting of maintenance strategies. Through Mahalanobis distance algorithm and covariance matrix modeling, it effectively suppresses interference from operating condition fluctuations, ensuring the stability of evaluation results in harsh environments such as high temperature, high corrosion, and high dust. Attached Figure Description

[0029] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0030] Figure 1 This is a flowchart of the AI-based device health assessment method in this embodiment.

[0031] Figure 2 This is a flowchart of the method for analyzing health characteristic parameters in this embodiment.

[0032] Figure 3 This is a flowchart of the method for analyzing health characteristic parameters in this embodiment.

[0033] Figure 4 This is a schematic diagram of the AI-based device health assessment device in this embodiment. Detailed Implementation

[0034] To more clearly illustrate the present invention, the following description, in conjunction with preferred embodiments and accompanying drawings, further explains the invention. Similar components in the drawings are indicated by the same reference numerals. Those skilled in the art should understand that the specific description below is illustrative rather than restrictive and should not be construed as limiting the scope of protection of the present invention.

[0035] It should be noted that although the terms first, second, third, etc., may be used in the embodiments of this application for description, these descriptions should not be limited to these terms. These terms are only used to distinguish the descriptions. For example, without departing from the scope of the embodiments of this application, first can also be referred to as second, and similarly, second can also be referred to as first.

[0036] Please see Figure 1 As shown, this is the AI-based device health assessment method in this embodiment, which includes:

[0037] Step S1: Real-time data collection of the guide wheel, chain, and slag / water data from the slag removal equipment.

[0038] The guide wheel data includes guide wheel torque, guide wheel rotation count, and grease life cycle count. The guide wheel rotation count is the sum of the number of rotations the guide wheel has made during the operation of the slag removal equipment. The grease life cycle count is the rated life cycle count of the grease used on the guide wheel. The guide wheel torque is in millimeters, and the guide wheel rotation count and grease life cycle count are in revolutions. The chain data includes chain scraper distance, design pitch, and chain load current. The chain scraper distance and design pitch are in millimeters, and the chain load current is in millimeters. The distance between the chain scrapers is amperes, the distance between the chain scrapers is the spacing between the chain and the adjacent scrapers, the design pitch is the preset spacing between the chain and the adjacent scrapers, the slag and water data includes liquid level height, slag density, flow rate and slag feed rate, the unit of liquid level height is meters, the unit of slag density is kilograms per cubic meter, the unit of flow rate is meters per second, and the unit of slag feed rate is cubic meters per second. The guide wheel data, chain data and slag and water data are collected and uploaded by sensors installed on the slag removal equipment.

[0039] Specifically, in step S1 of this embodiment, a multi-dimensional data acquisition network is constructed by integrating a guide wheel torque sensor, a chain displacement meter, and a multi-parameter probe for slag and water. This breakthrough allows for the simultaneous acquisition of real-time operating conditions of the mechanical transmission system and the fluid system, solving the blind spot problem of traditional single-sensor monitoring and providing a complete data foundation for subsequent multimodal health feature fusion.

[0040] Please continue reading. Figure 1 As shown, the AI-based device health assessment method further includes:

[0041] Step S2: Analyze the guide wheel torque fluctuation coefficient, chain pitch variation rate, and slag-water dynamic balance index based on the guide wheel data, chain data, and slag-water data of the slag removal equipment.

[0042] Specifically, in step S2 of this embodiment, the average torque of the guide wheel within 10 minutes is taken as the average torque value. Based on the guide wheel torque, the average torque value, the number of guide wheel rotations, and the number of grease life rotations, the guide wheel torque fluctuation coefficient is analyzed. The expression for the guide wheel torque fluctuation coefficient is: In the formula, TF(i) represents the guide wheel torque fluctuation coefficient, N(i) represents the number of guide wheel revolutions, T(i) represents the guide wheel torque, T1 represents the average torque, Nm represents the number of grease revolutions, i represents the time number, U1 represents the set of time numbers corresponding to 10 minutes, and NU1 represents the number of time numbers corresponding to 10 minutes. The time number is defined as a number to distinguish different data acquisition times; in this embodiment, the time number is used to represent data acquired in different seconds.

[0043] Specifically, in step S2 of this embodiment, the chain pitch variation rate is analyzed based on the chain scraper distance and the design pitch. The expression for the chain pitch variation rate is: PV(i)=Max(L(i|i∈U1))-Min(L(i|i∈U1))] / L1, where PV(i) represents the chain pitch variation rate, Max() represents the maximum value of the data extracted from the parentheses, Min() represents the minimum value of the data extracted from the parentheses, L(i) represents the chain scraper distance, and L1 represents the design pitch.

[0044] Specifically, in step S2 of this embodiment, the dynamic balance index of slag and water is analyzed based on the liquid level height, slurry density, flow rate and slag feed rate. The expression of the dynamic balance index of slag and water is: DB(i)=H(i)×ρ(i)×v(i) / Q(i), where DB(i) represents the dynamic balance index of slag and water, H(i) represents the liquid level height, ρ(i) represents the slurry density, v(i) represents the flow rate and Q(i) represents the slag feed rate.

[0045] Specifically, in step S2 of this embodiment, the impact of sealing performance degradation on transmission stability is quantified by integrating the number of life cycles of sealing grease and the real-time average torque value, thereby overcoming the problem of missed alarms caused by monitoring only the absolute torque value; the chain tension and wear degree are dynamically reflected based on the scraper distance range, replacing the traditional static pitch measurement method, thus improving the sensitivity to progressive mechanical deterioration; and the balance state of the slag-water system is diagnosed in real time through a nonlinear coupling model of liquid level, density, flow rate, and slag feed rate, avoiding incorrect adjustment of flushing volume due to a single liquid level alarm.

[0046] Please continue reading. Figure 1 As shown, the AI-based device health assessment method further includes:

[0047] Step S3 involves performing characteristic analysis on the guide wheel torque fluctuation coefficient, chain pitch variation rate, and slag-water dynamic balance index to obtain health characteristic parameters, including sealing performance index, wear acceleration, and risk deviation.

[0048] Please see Figure 2 As shown, it is a method for analyzing health characteristic parameters, including:

[0049] Step S31: Perform regression analysis on the guide wheel torque fluctuation coefficient to obtain the sealing performance index.

[0050] Specifically, in step S31 of this embodiment, guide wheel data satisfying TF(i)≥tf are extracted as sample guide wheel data, where tf represents the fluctuation coefficient extraction threshold. N(i) / Nm in the sample guide wheel data is used as the independent variable, and T(i) as the dependent variable, for regression analysis to obtain the regression equation for the guide wheel data. The regression equation is set as: y=a×x+b, where y represents the dependent variable, x represents the independent variable, a represents the first regression parameter, and b represents the second regression parameter. The sealing performance index is analyzed based on the guide wheel torque fluctuation coefficient, the first regression parameter, and the second regression parameter. The expression for the sealing performance index is: SE(i)=1 / [1+e -a×(TF(i)-b) ], where SE(i) represents the sealing performance index.

[0051] Specifically, in step S31 of this embodiment, abnormal samples are screened based on the fluctuation coefficient threshold, and the risk of seal failure is mapped using a logistic regression function to transform discrete torque fluctuations into continuous health indicators.

[0052] Please continue reading. Figure 2 As shown, the method for analyzing the health characteristic parameters further includes:

[0053] Step S32: Perform characteristic analysis on the chain pitch variation rate and chain load current to obtain the wear acceleration.

[0054] Specifically, in step S32 of this embodiment, the average value of the chain load current within 10 minutes is used as the average current parameter, and the wear acceleration is analyzed based on the average current parameter and the chain pitch variation rate. The expression for the wear acceleration is: WA(i)=d(PV(i)) / d(i)×Ravg, where WA(i) represents the wear acceleration, Ravg represents the average current parameter, and d(PV(i)) / d(i) represents the derivative of the chain pitch variation rate at time number i.

[0055] Specifically, in step S32 of this embodiment, a product model of the mean current parameter and the derivative of the pitch variation rate is introduced to dynamically characterize the chain wear rate and achieve early warning of sudden mechanical failures.

[0056] Please continue reading. Figure 2 As shown, the method for analyzing the health characteristic parameters further includes:

[0057] Step S33: Perform characteristic analysis on the slag-water dynamic balance index to obtain the risk deviation.

[0058] Specifically, in step S33 of this embodiment, the risk deviation is analyzed based on the slag-water dynamic balance index. The expression for the risk deviation is: RD(i)=|DB(i)-1|×d(DB(i)) / d(i), where RD(i) represents the risk deviation and d(DB(i)) / d(i) represents the derivative of the slag-water dynamic balance index at time number i.

[0059] Specifically, in step S33 of this embodiment, the overflow risk is quantified by combining the absolute value offset of the slag-water index with its trend derivative, thus solving the problem that traditional methods are not sensitive to gradual imbalance.

[0060] Please continue reading. Figure 1 As shown, the AI-based device health assessment method further includes:

[0061] Step S4: Construct a health assessment model based on health characteristic parameters to analyze the equipment health index.

[0062] Please see Figure 3 As shown, this is a method for constructing a health assessment model, including:

[0063] Step S41: Construct a health feature vector based on the health feature parameters and store the health feature vector.

[0064] Specifically, in step S41 of this embodiment, a health feature vector is constructed based on health feature parameters, and the health feature vector is set as...

[0065] Please continue reading. Figure 3 As shown, the method for constructing the health assessment model further includes:

[0066] Step S42: Analyze the health covariance matrix based on the stored health feature vectors.

[0067] Specifically, in step S42 of this embodiment, the health covariance matrix is ​​analyzed based on the stored health feature vectors from the past 30 days. The expression for the health covariance matrix is ​​as follows: In the formula, S represents the health covariance matrix, σ 2 () indicates the calculation of the variance of the data in parentheses, and σ(,) indicates the calculation of the covariance of the two sets of data in parentheses.

[0068] Please continue reading. Figure 3 As shown, the method for constructing the health assessment model further includes:

[0069] Step S43: Analyze the device health index based on the health covariance matrix and health eigenvectors.

[0070] Specifically, in step S43 of this embodiment, the device health index, which uses the Mahalanobis distance algorithm to analyze the health covariance matrix and health eigenvectors, is expressed as follows: In the formula, DH(i) represents the equipment health index. The mean vector representing the health feature vector.

[0071] Specifically, in step S4 of this embodiment, a health baseline domain is constructed using 30 days of historical data, and Mahalanobis distance is used to calculate the statistical deviation between real-time data and the baseline domain, so as to eliminate the interference of parameter dimension differences and correlations. The robustness to changes in complex working conditions is improved by using an adaptive covariance matrix, and the health index jump caused by transient abnormalities of individual parameters is avoided.

[0072] Please continue reading. Figure 1 As shown, the AI-based device health assessment method further includes:

[0073] Step S5: Develop an assessment decision tree based on the equipment health index.

[0074] Specifically, in step S5 of this embodiment, an evaluation decision tree is formulated based on the equipment health index. If the equipment health index is less than the first health threshold, the equipment health status is determined to be healthy; if the equipment health index is greater than or equal to the first health threshold and less than the second health threshold, the equipment health status is determined to be sub-healthy; if the equipment health index is greater than or equal to the second health threshold and less than the third health threshold, the equipment health status is determined to be high-risk; and if the equipment health index is greater than or equal to the third health threshold, the equipment health status is determined to be critically faulty.

[0075] Specifically, in this embodiment, the first health threshold is set to 1.5, the second health threshold is set to 2.5, and the third health threshold is set to 3.5. This embodiment does not impose specific limitations on the setting of the health thresholds, and those skilled in the art can set them freely, as long as they meet the requirements for determining the health status of the device.

[0076] Specifically, in step S5 of this embodiment, a four-level health status determination system is constructed, and a continuous threshold division based on the equipment health index is used to replace the traditional binary alarm mechanism. This enables progressive diagnosis of the equipment degradation process, dynamically correlates maintenance strategies, promotes the transformation from preventive maintenance to predictive maintenance, and reduces unplanned downtime.

[0077] Please continue reading. Figure 1 As shown, the AI-based device health assessment method further includes:

[0078] Step S6: Refine the evaluation decision tree based on the chain pitch variation rate, health characteristic parameters, and sludge-water dynamic balance index.

[0079] Specifically, in step S6 of this embodiment, the evaluation decision tree is refined based on the chain pitch variation rate, health characteristic parameters, equipment health index, and slag-water dynamic balance index. When the chain pitch variation rate is greater than the variation rate threshold and the wear acceleration is greater than the wear threshold, the equipment health status is determined to be chain abnormal. If the ratio of the standard deviation to the average value of the slag-water dynamic balance index within 10 minutes is greater than the fluctuation threshold, the equipment health status is determined to be overflow abnormal.

[0080] Specifically, in this embodiment, the mutation rate threshold is set to 0.05, the wear threshold is set to 0.8, and the fluctuation threshold is set to 0.15. This embodiment does not impose specific limitations on the settings of the mutation rate threshold, wear threshold, and fluctuation threshold. Those skilled in the art can set them freely, as long as they meet the requirements for refining the evaluation decision tree.

[0081] Specifically, in this embodiment, when the equipment is in a healthy state, normal inspection of the slag removal equipment is sufficient. When the equipment is in a sub-healthy state, the lubrication frequency of the slag removal equipment should be increased to ensure its normal operation. When the equipment is in a high-risk state, backup equipment should be used for operation and the equipment should be repaired within a short time. When the equipment is in a critical fault state, it should be stopped immediately for repair. When the equipment is in a chain abnormality state, the chain should be replaced to ensure the normal operation of the slag removal equipment. When the equipment is in an overflow abnormality state, the flushing water volume should be adjusted to ensure that the slag and water data are normal.

[0082] Specifically, in step S6 of this embodiment, by adding anomaly branch judgment based on key parameters, and combining the excessive pitch variation rate with the sharp increase in wear acceleration, it is possible to accurately distinguish between normal wear and abnormal jamming. By identifying short-term imbalance through the slag-water index fluctuation rate, it avoids misjudgment caused by the instantaneous change in slag feed rate, and realizes diagnosis from overall health assessment to subsystem fault tracing.

[0083] Please see Figure 4 As shown, this is an AI-based device health assessment device according to this embodiment, including:

[0084] The data acquisition unit is used to collect real-time data from the guide wheel, chain, and slag / water of the slag removal equipment.

[0085] The parameter analysis unit is used to analyze the guide wheel torque fluctuation coefficient, chain pitch variation rate, and slag-water dynamic balance index based on the guide wheel data, chain data, and slag-water data of the slag removal equipment.

[0086] The feature analysis unit is used to perform feature analysis on the guide wheel torque fluctuation coefficient, chain pitch variation rate and slag-water dynamic balance index to obtain health feature parameters;

[0087] The model building unit is used to build a health assessment model based on health characteristic parameters in order to analyze the equipment health index;

[0088] Decision-making unit, used to develop an assessment decision tree based on equipment health index;

[0089] The decision refinement unit is used to refine the evaluation decision tree based on the chain pitch variation rate, health characteristic parameters, and slag-water dynamic balance index.

[0090] This application also provides a computer-readable storage medium storing instructions that, when run on a computer, cause the computer to perform the AI-based device health assessment method as described in the above method embodiments.

[0091] It will be understood by those skilled in the art that all or some of the steps and systems in the methods disclosed above can be implemented as software, firmware, hardware, and suitable combinations thereof. Some or all of the physical components can be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit. Such software can be distributed on a computer-readable medium, which can include computer storage media (or non-transitory media) and communication media (or transient media). As is known to those skilled in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable programs, data structures, program modules, or other data). Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and is accessible to a computer. Furthermore, as is known to those skilled in the art, communication media typically contain computer-readable programs, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

[0092] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the implementation of the present invention. For those skilled in the art, other variations or modifications can be made based on the above description. It is impossible to exhaustively list all the implementation methods here. All obvious variations or modifications derived from the technical solutions of the present invention are still within the protection scope of the present invention.

Claims

1. An AI-based device health assessment method, characterized in that, include: Real-time data collection of guide wheel, chain, and slag / water data from the slag removal equipment; Based on the guide wheel data, chain data, and slag-water data of the slag removal equipment, the guide wheel torque fluctuation coefficient, chain pitch variation rate, and slag-water dynamic balance index were analyzed. Characteristic analysis was performed on the guide wheel torque fluctuation coefficient, chain pitch variation rate, and slag-water dynamic balance index to obtain health characteristic parameters; A health assessment model is constructed based on health characteristic parameters to analyze the equipment health index; Develop an assessment decision tree based on equipment health index; The evaluation decision tree is refined based on the chain pitch variation rate, health characteristic parameters, and sludge-water dynamic balance index.

2. The AI-based device health assessment method according to claim 1, characterized in that, The average torque of the guide wheel within 10 minutes is taken as the torque mean. The guide wheel torque fluctuation coefficient is analyzed based on the guide wheel torque, the average torque, the number of guide wheel turns, and the number of grease life turns. The chain pitch variation rate is analyzed based on the chain scraper distance and the design pitch. The slag-water dynamic balance index is analyzed based on the liquid level height, slurry density, flow rate, and slag feed rate.

3. The AI-based device health assessment method according to claim 2, characterized in that, The guide wheel data satisfying TF(i)≥tf are extracted as sample guide wheel data, where tf represents the fluctuation coefficient extraction threshold, TF(i) represents the guide wheel torque fluctuation coefficient, and i represents the time number. Regression analysis is performed on the sample guide wheel data to obtain the first regression parameter and the second regression parameter. The sealing performance index is analyzed based on the guide wheel torque fluctuation coefficient, the first regression parameter, and the second regression parameter. The expression for the sealing performance index is: SE(i)=1 / [1+e -a×(TF(i)-b) ], where SE(i) represents the sealing performance index.

4. The AI-based device health assessment method according to claim 3, characterized in that, The average value of the chain load current within 10 minutes is used as the average current parameter. The wear acceleration is analyzed based on the average current parameter and the chain pitch variation rate. The expression for the wear acceleration is: WA(i)=d(PV(i)) / d(i)×Ravg, where WA(i) represents the wear acceleration, Ravg represents the average current parameter, d(PV(i)) / d(i) represents the derivative of the chain pitch variation rate at time number i, and PV(i) represents the chain pitch variation rate.

5. The AI-based device health assessment method according to claim 4, characterized in that, The risk deviation is analyzed based on the slag-water dynamic balance index. The expression for the risk deviation is: RD(i)=|DB(i)-1|×d(DB(i)) / d(i), where RD(i) represents the risk deviation, d(DB(i)) / d(i) represents the derivative of the slag-water dynamic balance index at time number i, and DB(i) represents the slag-water dynamic balance index.

6. The AI-based device health assessment method according to claim 5, characterized in that, A health feature vector is constructed based on health feature parameters, and the health covariance matrix is ​​analyzed based on the health feature vectors stored over 30 days. The device health index of the health covariance matrix and health feature vectors is analyzed using the Mahalanobis distance algorithm.

7. The AI-based device health assessment method according to claim 6, characterized in that, An evaluation decision tree is developed based on the equipment health index. If the equipment health index is less than the first health threshold, the equipment health status is determined to be healthy. If the equipment health index is greater than or equal to the first health threshold and less than the second health threshold, the equipment health status is determined to be sub-healthy. If the equipment health index is greater than or equal to the second health threshold and less than the third health threshold, the equipment health status is determined to be a high-risk state. If the equipment health index is greater than or equal to the third health threshold, the equipment health status is determined to be a critical fault state.

8. The AI-based device health assessment method according to claim 7, characterized in that, The evaluation decision tree is refined based on the chain pitch variation rate, health characteristic parameters, equipment health index, and slag-water dynamic balance index. When the chain pitch variation rate is greater than the variation rate threshold and the wear acceleration is greater than the wear threshold, the equipment health status is determined to be chain abnormal. If the ratio of the standard deviation to the average value of the slag-water dynamic balance index within 10 minutes is greater than the fluctuation threshold, the equipment health status is determined to be overflow abnormal.

9. An AI-based device for assessing device health, applied to the AI-based device health assessment method as described in claims 1-8, characterized in that, include: The data acquisition unit is used to collect real-time data from the guide wheel, chain, and slag / water of the slag removal equipment. The parameter analysis unit is used to analyze the guide wheel torque fluctuation coefficient, chain pitch variation rate, and slag-water dynamic balance index based on the guide wheel data, chain data, and slag-water data of the slag removal equipment. The feature analysis unit is used to perform feature analysis on the guide wheel torque fluctuation coefficient, chain pitch variation rate and slag-water dynamic balance index to obtain health feature parameters; The model building unit is used to build a health assessment model based on health characteristic parameters in order to analyze the equipment health index; Decision-making unit, used to develop an assessment decision tree based on equipment health index; The decision refinement unit is used to refine the evaluation decision tree based on the chain pitch variation rate, health characteristic parameters, and slag-water dynamic balance index.

10. A storage medium, characterized in that, The device stores instructions that, when executed on a computer, cause the computer to perform the AI-based device health assessment method as described in any one of claims 1-8.

Images