Fault processing method, device and equipment for dry type slag remover of thermal power plant and medium

Abstract

The invention provides a fault processing method, device and equipment for a dry type slag remover of a thermal power plant and a medium. The method comprises the steps that multiple types of sensors are deployed for key fault points of the dry type slag remover to collect time sequence data in real time; preprocessing the collected data, including abnormal value removal and standardization, and missing value filling; inputting the preprocessed data into a CNN-LSTM model for training to capture spatial association and time sequence trends among different sensors, obtaining a trained fault prediction model according to a preset convergence condition, and outputting a fault type, a confidence coefficient and an abnormal parameter; and according to the fault type and the abnormal parameter output by the model, determining a corresponding fault level and sending out corresponding alarm information according to a preset reference threshold value of each sensor. According to the method, the fault type of the dry type slag remover can be quickly and accurately diagnosed, effective treatment measures are taken, the operation reliability of the slag remover is improved, the maintenance cost is reduced, and stable operation of a thermal power plant is guaranteed.

Description

Technical Field

[0001] This invention relates to the field of data analysis technology, and in particular to a method, apparatus, equipment and medium for troubleshooting dry slag removers in thermal power plants. Background Technology

[0002] Dry ash separators are crucial equipment in thermal power plant operations, used to process boiler ash. However, during long-term operation, dry ash separators are prone to various malfunctions, such as scraper chain failures, ash discharge gate malfunctions, and ash outlet blockages. These malfunctions not only affect the normal operation of the ash separator but may also lead to abnormal boiler operation and even impact the overall production efficiency and safety of the thermal power plant. Currently, most methods for handling dry ash separator malfunctions are relatively traditional, suffering from inaccurate fault diagnosis, long processing times, and high maintenance costs, failing to meet the demands of efficient and stable operation in thermal power plants. Therefore, a new method for handling dry ash separator malfunctions in thermal power plants is urgently needed to address the problems existing in current technologies.

[0003] Regular inspections by maintenance personnel are conducted, using visual inspection, auditory inspection, and tactile inspection to determine if the dry slag remover is malfunctioning. For example, they observe whether the scraper chain runs smoothly, listen for abnormal noises during operation, and touch the equipment casing to check for excessive temperature. When an abnormality is found, they analyze the cause of the malfunction based on their experience in handling similar problems. For a loose scraper chain, they manually adjust the tensioning device using tools; if the chain is broken, they manually remove and replace the damaged section. For slag discharge gate malfunctions, if it cannot be opened or closed, they manually check whether the mechanical connections are stuck or deformed, manually clearing any obstructions or repairing deformed parts. When the slag outlet is blocked, it is usually cleared manually using tools such as shovels.

[0004] Relying on personal experience, the varying skill levels and judgment abilities of different personnel make it difficult to guarantee the accuracy of fault diagnosis. Furthermore, manual inspections have time intervals, making it difficult to detect faults in a timely manner, allowing them to worsen unnoticed. When handling faults, the lack of precise fault location leads to somewhat aimless maintenance work, resulting in prolonged processing times and disruptions to the normal production of thermal power plants. Summary of the Invention

[0005] To overcome the problems existing in the related technologies, this disclosure provides a method, device, equipment and medium for troubleshooting dry ash removal machines in thermal power plants, so as to solve the technical problems in the related technologies.

[0006] This specification provides one or more embodiments of a method for troubleshooting a dry ash removal machine in a thermal power plant, including the following steps: By deploying multiple types of sensors to collect real-time time-series data at key fault points of the dry slag remover, including using vibration sensors to detect the chain's running status and temperature sensors to determine temperature changes; installing pressure sensors on the drive unit of the slag discharge gate to monitor the force applied when the gate opens and closes; and installing high-temperature resistant flow meters near the slag outlet to detect whether the outlet is blocked in real time. The collected data is preprocessed, including outlier removal and standardization, and missing value imputation. The preprocessed data is input into the CNN-LSTM model for training, which is used to capture the spatial correlation and temporal trend between different sensors. The trained fault prediction model is obtained according to the preset convergence conditions, and the fault type, confidence level and abnormal parameters are output. Based on the fault type and abnormal parameters output by the model, and according to the preset reference thresholds of each sensor, the corresponding fault level is determined and the corresponding alarm information is issued.

[0007] This specification provides one or more embodiments of a fault handling device for a dry ash removal machine in a thermal power plant, comprising: The data acquisition module is used to collect real-time time-series data by deploying multiple types of sensors at key fault points of the dry slag remover. This includes detecting the running status of the chain through vibration sensors, determining temperature changes by collecting data from temperature sensors, installing pressure sensors on the drive device of the slag discharge gate to monitor the force when the slag discharge gate is opened and closed, and installing a high-temperature resistant flow meter near the slag outlet to detect whether the slag outlet is blocked in real time. The data preprocessing module is used to preprocess the collected data, including outlier removal and standardization, and missing value imputation. The prediction module is used to input the preprocessed data into the CNN-LSTM model for training. It is used to capture the spatial correlation and temporal trend between different sensors. According to the preset convergence conditions, the trained fault prediction model is obtained and the fault type, confidence level and abnormal parameters are output. The fault level judgment and alarm module is used to determine the corresponding fault level and issue corresponding alarm information based on the fault type and abnormal parameters output by the model and the preset reference thresholds of each sensor.

[0008] This specification provides one or more embodiments of a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the fault handling method for a dry slag remover in a thermal power plant as described above.

[0009] This specification provides one or more embodiments of a storage medium, a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for troubleshooting dry slag removers in thermal power plants.

[0010] This disclosure provides a method, apparatus, equipment, and medium for troubleshooting dry ash removal machines in thermal power plants. Its advantages lie in its real-time, multi-parameter data acquisition of the dry ash removal machine by deploying various types of sensors (temperature, vibration, pressure, etc.) at key locations such as the scraper chain, ash discharge gate, and ash outlet. This overcomes the subjective limitations of manual experience-based judgment and the inadequacy of simple instruments that can only detect a single parameter, providing a more comprehensive and accurate reflection of the equipment's operating status. Furthermore, a diagnostic model trained on artificial intelligence algorithms and extensive historical fault data allows for in-depth analysis of the sensor-collected operating parameters. The model can intelligently determine whether a fault has occurred and its specific type based on data change characteristics. Compared to manual diagnosis relying on experience and simple instrument-based threshold judgment, this effectively avoids inaccurate diagnosis, false alarms, and missed alarms caused by individual skill differences and crude threshold settings, greatly improving diagnostic efficiency and accuracy. Finally, from fault monitoring to alarm and handling suggestions, a highly integrated automated and information-based process is formed. Once a fault is detected, the system automatically controls the ash removal machine to stop, issues audible and visual alarms and SMS alerts, and presents detailed fault information and handling suggestions on the monitoring interface. This changes the delays and blind spots of traditional manual fault handling, enabling rapid response and precise handling, significantly reducing downtime and improving equipment reliability. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in one or more embodiments of this specification or in the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments recorded in this specification. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 A flowchart illustrating a fault handling method for a dry ash removal machine in a thermal power plant, provided for one or more embodiments of this specification; Figure 2 A block diagram of a fault handling device for a dry slag remover in a thermal power plant, provided for one or more embodiments of this specification; Figure 3 This is a schematic diagram of the structure of a computer device provided for one or more embodiments of this specification. Detailed Implementation

[0013] To enable those skilled in the art to better understand the technical solutions in one or more embodiments of this specification, the technical solutions in one or more embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this specification, and not all of the embodiments. Based on one or more embodiments of this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of this invention.

[0014] The present invention will now be described in detail with reference to specific embodiments and accompanying drawings.

[0015] Method Implementation Examples According to embodiments of the present invention, a method for troubleshooting dry ash removal machines in thermal power plants is provided, such as... Figure 1 The diagram shown is a flowchart of a fault handling method for a dry ash removal machine in a thermal power plant according to this embodiment. The fault handling method for a dry ash removal machine in a thermal power plant according to this embodiment includes the following steps: Step S1 involves deploying multiple types of sensors to collect real-time time-series data at key fault points of the dry slag remover. This includes using vibration sensors to detect the chain's operating status, temperature sensors to determine temperature changes, installing pressure sensors on the slag discharge gate's drive unit to monitor the force applied when the gate opens and closes, and installing high-temperature resistant flow meters near the slag outlet to detect blockages in real time. The collected data is stored in the format of {timestamp, vibration value, temperature value, pressure value, current value, flow rate value, slag temperature value, fault label (normal / fault type)}.

[0016] In one specific embodiment, the deployment of multiple types of sensors at key fault points of the dry slag remover to collect real-time time-series data includes: vibration sensors and temperature sensors installed at preset intervals on the scraper chain to collect vibration acceleration and chain surface temperature; pressure sensors and current sensors on the motor power line deployed at the cylinder of the slag discharge gate drive device to collect cylinder pressure and motor current; and high-temperature flow meters and infrared temperature sensors installed on the slag discharge pipe and the inner wall of the slag discharge port to collect slag discharge flow rate and outlet slag temperature, respectively.

[0017] Step S2 involves preprocessing the collected data, including outlier removal and standardization, and missing value imputation.

[0018] In one embodiment, outlier removal and standardization are performed as follows: Step 11, using IQR The interquartile range (ICM) method is used for outlier removal, and the calculation formula is as follows: Quartiles:Q1 (25th percentile) Q3 (75th percentile); Interquartile range: IQR = Q3 - Q1 ; Calculation of outlier range: [ Q1-1.5IQR , Q3+1.5IQR ].

[0019] Step 12: Use Z-score standardization to eliminate the influence of dimensions, as shown in the following formula: ; in, x For raw sensor data, μ This is the mean of the parameter. σ Standard deviation x norm This is the standardized data.

[0020] In this embodiment, due to sensor failure or transmission interruption causing data loss (typically accounting for 5%-15%), a variational autoencoder (VAE) is used to fill in the data, as detailed below: VAE structure encoder Map the known data to the latent variable z, decoder Generate imputed data, with the loss function being: ; in, The prior distribution of the latent variables is KL divergence constrained to approximate a normal distribution; the data missing rate after training is ≤0.15%.

[0021] Step S3: Input the preprocessed data into the CNN-LSTM model for training. This model is used to capture the spatial correlation and temporal trend between different sensors. The trained fault prediction model is obtained according to the preset convergence conditions, and the fault type, confidence level and abnormal parameters are output.

[0022] In this embodiment, the CNN-LSTM model includes convolutional layer 1, pooling layer 1, convolutional layer 2, pooling layer 2, flattening layer, LSTM layer, and fully connected layer. Convolutional layer 1 is used to extract spatial correlation features between sensors, pooling layer 1 reduces dimensionality and retains key spatial features, convolutional layer 2 enhances spatial feature extraction, flattening layer converts spatial features into vectors, LSTM layer is used to capture temporal features, and fully connected layer is used for feature fusion and fault type probability output.

[0023] The formula for feature extraction from convolutional layers is: ; The temporal features of the LSTM layer are calculated as follows: Input Gate: ; Forgotten Gate: ; Output gate:

[0024] Cell state: ; Hidden state: .

[0025] In this embodiment, the CNN-LSTM model optimizer uses the AdamW optimizer with a learning rate of 0.001; the loss function is cross-entropy loss; the number of training iterations is set to 80 rounds; and the batch size is 64.

[0026] Step S4: Based on the fault type and abnormal parameters output by the model, and according to the preset reference thresholds of each sensor, determine the corresponding fault level and issue the corresponding alarm information.

[0027] In this example, the alarm information includes the fault time, fault location, fault type, confidence level, and key abnormal parameters; and the fault alarm level is determined based on the confidence level setting. When the confidence level is ≥90%, the alarm level is emergency, and a red audible and visual alarm is activated. When the confidence level is ≥70%-89%, the alarm level is critical, and a yellow audible and visual alarm is triggered. When the confidence level is ≥50%-69%, the alarm level is prompt, and the system will display a pop-up alarm.

[0028] This embodiment also includes the step of: performing fault handling according to the fault type. For scraper chain failures, if the diagnosis is abnormal vibration but the temperature is normal, it indicates that the chain is loose. The system will automatically control the slag remover to reduce its speed to 50% and push the handling suggestion "adjust the tensioning device". After the maintenance personnel operate on-site, the system will monitor the vibration value in real time. After it returns to normal, it will automatically increase the speed. If the vibration value rises sharply and the temperature drops sharply, it is determined that the chain is broken. The system will immediately stop the machine, shut down the drive motor, and push the handling suggestion "remove the broken section and replace it with a chain link of the same model". After the handling is completed, the machine will be manually started for trial operation.

[0029] For scraper chain malfunctions, if the current is abnormal but the pressure is normal, it indicates a control circuit fuse failure. The system will automatically cut off the power supply to the slag discharge door drive. The recommended troubleshooting step is to "check the control circuit fuse and replace the damaged component." If the pressure is abnormal but the current is normal, it indicates a slag discharge door malfunction. The system will stop the slag discharge door from operating. The recommended troubleshooting step is to "clean the door of any obstructions and check the cylinder seals."

[0030] If the flow rate drops by 10%-30% and the slag temperature rises, it indicates a slight blockage at the slag outlet. The system will automatically activate the high-pressure air purging device and monitor the flow rate in real time. Purging will stop once the flow rate returns to normal. If the flow rate drops by ≥30% or the slag temperature rises sharply, it indicates a severe blockage at the slag outlet. The system will then reduce its speed to 30%, activate the mechanical unblocking device, and if the flow rate does not recover after a preset period, the system will automatically shut down and send a message to "manually clean the slag outlet".

[0031] The method provided in this embodiment has the following beneficial effects: (1) Multi-dimensional real-time monitoring system: By deploying various types of sensors such as temperature, vibration, and pressure in key parts such as scraper chain, slag discharge gate, and slag outlet, real-time data acquisition of multiple parameters of the dry slag remover can be achieved. This is different from the subjective limitations of human experience judgment and the inadequacy of simple instruments that can only detect a single parameter, and can more comprehensively and accurately reflect the operating status of the equipment.

[0032] (2) Intelligent Fault Diagnosis Model: Based on artificial intelligence algorithms and trained with a large amount of historical fault data, the diagnostic model can perform in-depth analysis of the operating parameters collected by sensors. The model can intelligently determine whether a fault has occurred and the specific type of fault based on the characteristics of data changes. Compared with manual diagnosis relying on experience and simple instruments relying on threshold judgment, it can effectively avoid the problems of inaccurate diagnosis, false alarms and missed alarms caused by differences in individual ability and rough threshold settings, and greatly improve the efficiency and accuracy of diagnosis.

[0033] (3) Integrated Automation and Information Processing: From fault monitoring to alarm and then to providing handling suggestions, a highly integrated automation and information process has been formed. Once a fault is detected, the system automatically controls the slag remover to stop, issues audible and visual alarms and SMS alerts, and displays detailed fault information and handling suggestions on the monitoring interface. This changes the delay and blindness of traditional manual fault handling, achieving rapid response and accurate handling, significantly reducing downtime and improving equipment reliability.

[0034] Device Examples According to embodiments of the present invention, a fault handling device for a dry ash removal machine in a thermal power plant is provided, such as... Figure 2 The diagram shown is a block diagram of a fault handling device for a dry ash removal machine in a thermal power plant according to this embodiment. The fault handling device for a dry ash removal machine in a thermal power plant according to this embodiment includes: The data acquisition module 10 is used to collect real-time time-series data by deploying multiple types of sensors at key fault points of the dry slag remover. This includes detecting the running status of the chain through vibration sensors, determining temperature changes by collecting data from temperature sensors, installing pressure sensors on the drive device of the slag discharge gate to monitor the force when the slag discharge gate is opened and closed, and installing a high-temperature resistant flow meter near the slag outlet to detect whether the slag outlet is blocked in real time. The collected data is stored in the format of {timestamp, vibration value, temperature value, pressure value, current value, flow rate value, slag temperature value, fault label (normal / fault type)}.

[0035] The data preprocessing module 20 is used to preprocess the collected data, including outlier removal and standardization, and missing value imputation. The prediction module 30 is used to input the preprocessed data into the CNN-LSTM model for training. It is used to capture the spatial correlation and temporal trend between different sensors, obtain the trained fault prediction model according to the preset convergence conditions, and output the fault type, confidence level and abnormal parameters.

[0036] In this embodiment, the CNN-LSTM model includes convolutional layer 1, pooling layer 1, convolutional layer 2, pooling layer 2, flattening layer, LSTM layer, and fully connected layer. Convolutional layer 1 is used to extract spatial correlation features between sensors, pooling layer 1 reduces dimensionality and retains key spatial features, convolutional layer 2 enhances spatial feature extraction, flattening layer converts spatial features into vectors, LSTM layer is used to capture temporal features, and fully connected layer is used for feature fusion and fault type probability output.

[0037] The formula for feature extraction from convolutional layers is: ; The temporal features of the LSTM layer are calculated as follows: Input Gate: ; Forgotten Gate: ; Output gate:

[0038] Cell state: ; Hidden state: .

[0039] In this embodiment, the CNN-LSTM model optimizer uses the AdamW optimizer with a learning rate of 0.001; the loss function is cross-entropy loss; the number of training iterations is set to 80 rounds; and the batch size is 64.

[0040] The fault level judgment alarm module 40 is used to determine the corresponding fault level and issue corresponding alarm information based on the fault type and abnormal parameters output by the model and the preset reference thresholds of each sensor.

[0041] In this example, the alarm information includes the fault time, fault location, fault type, confidence level, and key abnormal parameters; and the confidence level setting determines the fault alarm level. When the confidence level is ≥90%, the alarm level is emergency, and a red audible and visual alarm is activated. When the confidence level is ≥70%-89%, the alarm level is critical, and a yellow audible and visual alarm is triggered. When the confidence level is ≥50%-69%, the alarm level is prompt, and the system will display a pop-up alarm.

[0042] The embodiments of the present invention are device embodiments corresponding to the above method embodiments. The specific operations of each module processing step can be understood with reference to the description of the method embodiments, and will not be repeated here.

[0043] like Figure 3 As shown, the present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the fault handling method for the dry slag remover in the thermal power plant described in the above embodiments.

[0044] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the fault handling method for dry slag remover in thermal power plants as described in the above embodiments.

[0045] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0046] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, for apparatus or system embodiments, since they are basically similar to method embodiments, the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments. The apparatus and system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0047] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. 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 or all of the technical features. These modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and the contents not described in detail in the specification of the present invention are well known to those skilled in the art.

Claims

1. A method for troubleshooting a dry ash removal machine in a thermal power plant, characterized in that, Includes the following steps: By deploying multiple types of sensors to collect real-time time-series data at key fault points of the dry slag remover, including using vibration sensors to detect the chain's running status and temperature sensors to determine temperature changes; installing pressure sensors on the drive unit of the slag discharge gate to monitor the force applied when the gate opens and closes; and installing high-temperature resistant flow meters near the slag outlet to detect whether the outlet is blocked in real time. The collected data is preprocessed, including outlier removal and standardization, and missing value imputation. The preprocessed data is input into the CNN-LSTM model for training, which is used to capture the spatial correlation and temporal trend between different sensors. The trained fault prediction model is obtained according to the preset convergence conditions, and the fault type, confidence level and abnormal parameters are output. Based on the fault type and abnormal parameters output by the model, and according to the preset reference thresholds of each sensor, the corresponding fault level is determined and the corresponding alarm information is issued.

2. The method for troubleshooting a dry ash removal machine in a thermal power plant as described in claim 1, characterized in that, The preprocessing of the collected data includes the following steps: use IQR Methods for outlier removal; Z-score standardization is used to eliminate the influence of dimensions; A variational autoencoder is used to complete the data filling.

3. The troubleshooting method for dry ash removal machines in thermal power plants as described in claim 1, characterized in that, The CNN-LSTM model includes convolutional layer 1, pooling layer 1, convolutional layer 2, pooling layer 2, flattening layer, LSTM layer, and fully connected layer. Convolutional layer 1 is used to extract spatial correlation features between sensors, pooling layer 1 reduces dimensionality and retains key spatial features, convolutional layer 2 enhances spatial feature extraction, flattening layer converts spatial features into vectors, LSTM layer is used to capture temporal features, and fully connected layer is used for feature fusion and fault type probability output. in, The formula for feature extraction from convolutional layers is: ； The temporal features of the LSTM layer are calculated as follows: Input Gate: ; Forgotten Gate: ; Output gate: Cell state: ; Hidden state: .

4. The method for troubleshooting a dry ash removal machine in a thermal power plant as described in claim 1, characterized in that, The alarm information includes the fault time, fault location, fault type, confidence level, and key abnormal parameters; and the fault alarm level is determined according to the confidence level setting, as follows: When the confidence level is ≥90%, the alarm level is emergency, and a red audible and visual alarm is activated. When the confidence level is ≥70%-89%, the alarm level is critical, and a yellow audible and visual alarm is triggered. When the confidence level is ≥50%-69%, the alarm level is prompt, and the system will display a pop-up alarm.

5. The troubleshooting method for dry ash removal machines in thermal power plants as described in claim 1, characterized in that, It also includes fault handling steps based on the fault type, as detailed below: For scraper chain failures, if the diagnosis is abnormal vibration but normal temperature, it indicates that the chain is loose. The system will automatically control the slag remover to reduce its speed to 50%. After on-site operation by maintenance personnel, the system will monitor the vibration value in real time and automatically increase the speed after it returns to normal. If the vibration value increases suddenly and the temperature drops suddenly, it is determined that the chain is broken, and the system will immediately stop the machine and shut down the drive motor. If the scraper chain malfunctions and the current is abnormal while the pressure is normal, it indicates a fault in the control circuit fuse, and the system will automatically cut off the power supply to the slag discharge gate drive. If the pressure is abnormal while the current is normal, it indicates a fault in the slag discharge gate, and the system will stop the slag discharge gate from operating. If the flow rate drops by 10%-30% and the slag temperature rises, it indicates a slight blockage at the slag outlet. The system will automatically start the high-pressure air purging device and monitor the flow rate in real time. The purging will stop once the flow rate returns to normal. If the flow rate drops by ≥30% or the slag temperature rises sharply, it indicates a severe blockage at the slag outlet. The system will then reduce its speed to 30%, start the mechanical unblocking device, and automatically shut down the system if the flow rate does not recover after a preset time period.

6. A troubleshooting device for a dry ash removal machine in a thermal power plant, characterized in that, include: The data acquisition module is used to collect real-time time-series data by deploying multiple types of sensors at key fault points of the dry slag remover. This includes detecting the running status of the chain through vibration sensors, determining temperature changes by collecting data from temperature sensors, installing pressure sensors on the drive device of the slag discharge gate to monitor the force when the slag discharge gate is opened and closed, and installing a high-temperature resistant flow meter near the slag outlet to detect whether the slag outlet is blocked in real time. The data preprocessing module is used to preprocess the collected data, including outlier removal and standardization, and missing value imputation. The prediction module is used to input the preprocessed data into the CNN-LSTM model for training. It is used to capture the spatial correlation and temporal trend between different sensors. According to the preset convergence conditions, the trained fault prediction model is obtained and the fault type, confidence level and abnormal parameters are output. The fault level judgment and alarm module is used to determine the corresponding fault level and issue corresponding alarm information based on the fault type and abnormal parameters output by the model and the preset reference thresholds of each sensor.

7. The fault handling device for dry ash removal machines in thermal power plants as described in claim 6, characterized in that, The CNN-LSTM model includes convolutional layer 1, pooling layer 1, convolutional layer 2, pooling layer 2, flattening layer, LSTM layer, and fully connected layer. Convolutional layer 1 is used to extract spatial correlation features between sensors, pooling layer 1 reduces dimensionality and retains key spatial features, convolutional layer 2 enhances spatial feature extraction, flattening layer converts spatial features into vectors, LSTM layer is used to capture temporal features, and fully connected layer is used for feature fusion and fault type probability output. The formula for feature extraction from convolutional layers is: ； The temporal features of the LSTM layer are calculated as follows: Input Gate: ; Forgotten Gate: ; Output gate: Cell state: ; Hidden state: .

8. The fault handling device for dry ash removal machines in thermal power plants as described in claim 6, characterized in that, The alarm information includes the fault time, fault location, fault type, confidence level, and key abnormal parameters; and the confidence level setting determines the fault alarm level. When the confidence level is ≥90%, the alarm level is emergency, and a red audible and visual alarm is activated. When the confidence level is ≥70%-89%, the alarm level is critical, and a yellow audible and visual alarm is triggered. When the confidence level is ≥50%-69%, the alarm level is prompt, and the system will display a pop-up alarm.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the fault handling method for dry slag removal machine in thermal power plants as described in any one of claims 1 to 5.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the fault handling method for dry slag removal machines in thermal power plants as described in any one of claims 1 to 5.

Images