Intelligent anti-deviation self-adaptive control method of coal coke transfer belt conveyor

Abstract

This invention relates to the field of belt conveyor technology, specifically to an intelligent anti-deviation adaptive control method for a coal and coke conveyor belt, comprising: an intelligent anti-deviation adaptive control method for a coal and coke conveyor belt, characterized by the following steps: Step 1: Multi-dimensional data acquisition (real-time sensing), S1.1 Deviation detection: One set of laser displacement sensor arrays (detection accuracy ±1mm) are arranged at the head, middle, and tail of the belt conveyor, and an industrial camera is deployed for visual recognition to collect the belt lateral offset (0~100mm), offset direction (left / right), and offset rate (mm / s) in real time; This invention adopts multi-dimensional detection using laser array + visual recognition, with a deviation detection response time ≤0.5s, offset detection accuracy ±1mm, and can trigger control when the offset is ≥5mm, completely avoiding belt friction with the frame, reducing the belt edge wear rate to ≤0.05mm / day, reducing the tear accident rate to 0, and reducing the downtime maintenance rate to ≤1 time / month.

Description

Technical Field

[0001] This invention relates to the field of belt conveyor technology, specifically to an intelligent anti-deviation adaptive control method for a coal and coke conveyor belt. Background Technology

[0002] Coal and coke conveyor belts are core equipment for material transportation in the coal and coking industries, and their operational stability directly affects production continuity. However, existing belt conveyor anti-deviation technology has significant bottlenecks:

[0003] Delayed belt misalignment detection and passive control: Traditional methods use manual inspection or single-point limit switches to detect belt misalignment, with a detection lag of ≥5s. Alarms can only be triggered when the belt offset is ≥50mm. At this time, the belt has already rubbed against the frame, which can easily cause belt edge wear (wear rate ≥0.5mm / day) and tearing, resulting in a downtime maintenance rate of ≥1 time / week.

[0004] The control method is crude and has poor adaptability: it mostly uses fixed-angle adjustment rollers or manual adjustment drums, which cannot be dynamically adjusted according to the direction of deviation, the amount of deviation, the material load (0~100% of the rated load), and the belt speed (1~3m / s). The adjustment accuracy is ≤±20mm, and it is easy to "over-adjust" or "under-adjust", with a secondary deviation rate ≥30%.

[0005] Lacking adaptive learning capabilities and insufficient adaptability to operating conditions: When operating conditions change during coal and coke transfer, such as uneven material distribution (unilateral load rate ≥20%), belt aging (elastic modulus attenuation ≥10%), and high environmental dust (dust concentration ≥50mg / m³), the anti-deviation parameters cannot be optimized autonomously, and the probability of control failure is ≥25%.

[0006] The disconnect between dust removal and deviation prevention exacerbates the risk of deviation: Traditional dust removal devices are mostly fixed air outlets, and the impact of airflow causes the material to be further misaligned. In addition, dust covers the detection sensors, reducing the accuracy of deviation detection to ≤70%, forming a vicious cycle of "dust removal exacerbating deviation, and deviation affecting detection".

[0007] Isolated data and lack of predictive capability: No correlation model has been established between belt misalignment data and operating conditions, making it impossible to predict belt misalignment trends. Only post-mortem remediation is possible. Sudden misalignment can result in material spillage losses of ≥0.5 tons per incident, increasing equipment maintenance costs by more than 40%.

[0008] Existing technologies only focus on "single-point detection + passive deviation adjustment" and have not formed an intelligent system of "multi-dimensional perception - dynamic adaptive control - dust removal and deviation prevention coordination - trend prediction". Therefore, they cannot adapt to the complex and variable working conditions of coal and coke transfer. To this end, an intelligent deviation prevention adaptive control method for coal and coke transfer belt conveyors is proposed. Summary of the Invention

[0009] In view of this, the present invention provides an intelligent anti-deviation adaptive control method for coal and coke conveyor belts to solve or alleviate the technical problems existing in the prior art, and at least provides a beneficial option.

[0010] The technical solution of this invention is implemented as follows: an intelligent anti-deviation adaptive control method for a coal and coke conveyor belt, characterized by comprising the following steps:

[0011] Step 1: Multi-dimensional data collection (real-time sensing)

[0012] S1.1 Misalignment Detection: One set of laser displacement sensor arrays (detection accuracy ±1mm) is arranged at the head, middle and tail of the belt conveyor. At the same time, an industrial camera is deployed for visual recognition to collect the belt lateral offset (0~100mm), offset direction (left / right offset) and offset rate (mm / s) in real time.

[0013] S1.2 Operating Condition Data Acquisition: The real-time load of the belt (0~100% of the rated load) is collected by the load sensor, the belt running speed (1~3m / s) is collected by the speed sensor, the tilt angle of the idler roller is collected by the tilt sensor, and the ambient dust concentration (0~100mg / m³) is collected by the dust sensor.

[0014] S1.3 Status Acquisition: Acquire status parameters such as the current angle, drive motor current, and belt tension (0~50kN) of the alignment mechanism (alignment roller, alignment drum);

[0015] Step 2: Data Preprocessing and Intelligent Analysis

[0016] S2.1 Data Cleaning: Remove sensor outliers (such as abrupt data changes caused by dust obstruction), and fuse and calibrate laser and vision inspection data to ensure that the offset detection error is ≤ ±1mm;

[0017] S2.2 Deviation Level Determination: Deviation levels are determined based on the amount of offset.

[0018] Level 1 (early warning): 5-20mm, initiate pre-regulation;

[0019] Level 2 (Mild): 20-50mm, initiate normal control;

[0020] Level 3 (Severe): 50mm, activate emergency control and alarm;

[0021] S2.3 Intelligent Modeling and Analysis: Based on machine learning algorithms (random forest), a correlation model of "offset - load - speed - tension - adjustment parameters" is established, and the optimal adjustment strategy (adjustment roller angle, correction roller displacement, tension adjustment value) is output.

[0022] Step 3: Adaptive Adjustment Execution

[0023] S3.1 Tiered Regulation:

[0024] Level 1 deviation: Adjust the tilt angle of the adjustment roller (adjustment range 0~15°, adjustment step 0.5°), and simultaneously adjust the angle of the adaptive dust removal air outlet to avoid uneven airflow.

[0025] Secondary deviation: Based on the adjustment of the deviation idler, fine-tune the lateral displacement of the correction roller (adjustment range 0~50mm, adjustment step 1mm) to adapt to load / speed changes;

[0026] Level 3 belt misalignment: immediately reduce belt speed to ≤1m / s, and simultaneously activate the double-sided adjustment rollers and correction drum for linkage adjustment, with adaptive tension compensation (adjustment range ±5kN).

[0027] S3.2 Real-time feedback: The offset is collected every 0.1s during the adjustment process. If the offset does not decrease after adjustment, the adjustment parameters are automatically iterated and optimized until the offset is ≤5mm.

[0028] Step 4: Coordinated Control of Dust Removal and Deviation Prevention

[0029] S4.1 Adaptive Dust Removal: Adjust the angle (0~30°) and air volume (500~2000m³ / h) of the dust removal vent according to the direction / amount of belt offset, so that the airflow is perpendicular to the belt surface and avoids the airflow causing material to be unbalanced;

[0030] S4.2 Sensor Dust Prevention: Adjust the dust removal air vents to perform directional blowing on the laser / vision sensor area, with a blowing air volume of 50-100 m³ / h, to ensure sensor cleanliness and detection accuracy ≥99%;

[0031] Step 5: Self-learning and trend prediction

[0032] S5.1 Self-learning optimization: Daily review of the day's deviation data, control effect, and operating parameters, iteratively update the machine learning model, and optimize the adjustment parameters under different operating conditions;

[0033] S5.2 Trend Prediction: Based on historical data and real-time operating conditions, predict the belt deviation trend in the next 5 to 10 seconds. If the predicted deviation is ≥10mm, start pre-adjustment in advance to avoid further deviation.

[0034] Step 6: Fault Alarm and Data Traceability

[0035] S6.1 Alarm Trigger: When the third-level deviation lasts for ≥3 seconds, the deviation adjustment mechanism fails (abnormal motor current), or the sensor fails, an audible and visual alarm is triggered and pushed to the central control system;

[0036] S6.2 Data Traceability: Records all deviation events, control processes, and operating parameters, with a retention period of ≥1 year, supporting fault tracing and process optimization.

[0037] More preferably, the method is applicable to coal and coke conveyor belts with a bandwidth of 500-2000mm, a belt speed of 1-3m / s, and a rated load of 0-500t / h, with a deviation control accuracy of ±5mm and a control response time of ≤0.5s.

[0038] More preferably, in step S1.1, the detection frequency of the laser displacement sensor array is 100Hz, the sampling frame rate of the industrial camera is 30fps, and the offset detection accuracy after data fusion is ≤±1mm.

[0039] More preferably, in step S2.3, the training samples of the machine learning model are ≥100,000 sets, covering all working conditions of load 0-100%, speed 1-3m / s, and offset 0-100mm, and the accuracy of the model output adjustment parameters is ≥98%.

[0040] More preferably, in step S3.1, the adjustment of the alignment roller is driven by a servo motor with an angle adjustment accuracy of ±0.1°, and the displacement of the alignment roller is driven by a ball screw with a displacement accuracy of ±0.1mm.

[0041] In a further preferred embodiment, in step S4.1, the adaptive dust removal vent uses an electric actuator to adjust the angle, with an adjustment range of 0~30° and an adjustment accuracy of ±0.5°, and the air volume is controlled by a variable frequency fan, with a control range of 500~2000m³ / h.

[0042] More preferably, the iteration cycle of self-learning optimization in step S5.1 is 24 hours, and the model adjustment success rate is improved by ≥0.5% after each iteration, which is suitable for the working condition of belt aging (elastic modulus decay of 0 to 20%).

[0043] More preferably, the trend prediction in step S5.2 adopts an LSTM neural network model, and the input parameters include real-time offset, offset rate, load, speed, and tension force, with a prediction accuracy of ≥95%.

[0044] More preferably, the alarm information in step S6.1 includes the deviation level, the location of occurrence, the current operating condition, and the suggested control scheme, and the response time to push it to the central control system is ≤1s.

[0045] The embodiments of the present invention have the following advantages due to the adoption of the above technical solutions:

[0046] I. This invention employs multi-dimensional detection using a laser array and visual recognition. The belt misalignment detection response time is ≤0.5s, and the offset detection accuracy is ±1mm. Adjustment can be triggered when the offset is ≥5mm, completely avoiding belt friction with the frame. The belt edge wear rate is reduced to ≤0.05mm / day, the tearing accident rate is reduced to 0, and the downtime maintenance rate is ≤1 time / month. The adjustment mechanism parameters are dynamically adjusted based on the misalignment direction, offset, load, and speed, with an adjustment accuracy of ±5mm and an over / under adjustment rate of ≤1%. It is suitable for all working conditions from 0 to 100% rated load and belt speed from 1 to 3m / s, with an adjustment success rate of ≥99%.

[0047] II. The present invention enables the adaptive dust removal vent to dynamically adjust according to belt misalignment, improving airflow uniformity by 80%, reducing dust concentration to ≤10mg / m³, maintaining sensor detection accuracy at ≥99%, and avoiding material misalignment caused by airflow, reducing the misalignment rate to ≤5%. By optimizing control parameters through machine learning, it adapts to working conditions such as belt aging and material distribution changes, predicting misalignment trend with an accuracy of ≥95%, and initiating pre-control 5-10 seconds in advance, reducing the sudden misalignment rate to ≤2%.

[0048] Third, the belt replacement cycle of this invention is extended by 2 times, material spillage loss is reduced by 95%, and equipment maintenance costs are reduced by more than 60%; intelligent control enables unattended operation, manual inspection costs are reduced by 80%, the central control system displays deviation data, control process, and equipment status in real time, forms a coal and coke transfer belt conveyor operation ledger, the fault tracing time is shortened from ≥2h to ≤5min, and management efficiency is improved by 70%.

[0049] The above overview is for illustrative purposes only and is not intended to be limiting in any way. In addition to the illustrative aspects, embodiments, and features described above, further aspects, embodiments, and features of the invention will become readily apparent from the accompanying drawings and the following detailed description. Attached Figure Description

[0050] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0051] Figure 1 This is a process flow diagram of the present invention. Detailed Implementation

[0052] In the following description, only certain exemplary embodiments are briefly described. As those skilled in the art will recognize, the described embodiments can be modified in various ways without departing from the spirit or scope of the invention. Therefore, the drawings and description are considered to be exemplary in nature and not restrictive.

[0053] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0054] like Figure 1 As shown in the figure, this invention provides an intelligent anti-deviation adaptive control method for a coal and coke conveyor belt, characterized by the following steps:

[0055] Step 1: Multi-dimensional data collection (real-time sensing)

[0056] S1.1 Misalignment Detection: One set of laser displacement sensor arrays (detection accuracy ±1mm) is arranged at the head, middle and tail of the belt conveyor. At the same time, an industrial camera is deployed for visual recognition to collect the belt lateral offset (0~100mm), offset direction (left / right offset) and offset rate (mm / s) in real time.

[0057] S1.2 Operating Condition Data Acquisition: The real-time load of the belt (0~100% of the rated load) is collected by the load sensor, the belt running speed (1~3m / s) is collected by the speed sensor, the tilt angle of the idler roller is collected by the tilt sensor, and the ambient dust concentration (0~100mg / m³) is collected by the dust sensor.

[0058] S1.3 Status Acquisition: Acquire status parameters such as the current angle, drive motor current, and belt tension (0~50kN) of the alignment mechanism (alignment roller, alignment drum);

[0059] Step 2: Data Preprocessing and Intelligent Analysis

[0060] S2.1 Data Cleaning: Remove sensor outliers (such as abrupt data changes caused by dust obstruction), and fuse and calibrate laser and vision inspection data to ensure that the offset detection error is ≤ ±1mm;

[0061] S2.2 Deviation Level Determination: Deviation levels are determined based on the amount of offset.

[0062] Level 1 (early warning): 5-20mm, initiate pre-regulation;

[0063] Level 2 (Mild): 20-50mm, initiate normal control;

[0064] Level 3 (Severe): 50mm, activate emergency control and alarm;

[0065] S2.3 Intelligent Modeling and Analysis: Based on machine learning algorithms (random forest), a correlation model of "offset - load - speed - tension - adjustment parameters" is established, and the optimal adjustment strategy (adjustment roller angle, correction roller displacement, tension adjustment value) is output.

[0066] Step 3: Adaptive Adjustment Execution

[0067] S3.1 Tiered Regulation:

[0068] Level 1 deviation: Adjust the tilt angle of the adjustment roller (adjustment range 0~15°, adjustment step 0.5°), and simultaneously adjust the angle of the adaptive dust removal air outlet to avoid uneven airflow.

[0069] Secondary deviation: Based on the adjustment of the deviation idler, fine-tune the lateral displacement of the correction roller (adjustment range 0~50mm, adjustment step 1mm) to adapt to load / speed changes;

[0070] Level 3 belt misalignment: immediately reduce belt speed to ≤1m / s, and simultaneously activate the double-sided adjustment rollers and correction drum for linkage adjustment, with adaptive tension compensation (adjustment range ±5kN).

[0071] S3.2 Real-time feedback: The offset is collected every 0.1s during the adjustment process. If the offset does not decrease after adjustment, the adjustment parameters are automatically iterated and optimized until the offset is ≤5mm.

[0072] Step 4: Coordinated Control of Dust Removal and Deviation Prevention

[0073] S4.1 Adaptive Dust Removal: Adjust the angle (0~30°) and air volume (500~2000m³ / h) of the dust removal vent according to the direction / amount of belt offset, so that the airflow is perpendicular to the belt surface and avoids the airflow causing material to be unbalanced;

[0074] S4.2 Sensor Dust Prevention: Adjust the dust removal air vents to perform directional blowing on the laser / vision sensor area, with a blowing air volume of 50-100 m³ / h, to ensure sensor cleanliness and detection accuracy ≥99%;

[0075] Step 5: Self-learning and trend prediction

[0076] S5.1 Self-learning optimization: Daily review of the day's deviation data, control effect, and operating parameters, iteratively update the machine learning model, and optimize the adjustment parameters under different operating conditions;

[0077] S5.2 Trend Prediction: Based on historical data and real-time operating conditions, predict the belt deviation trend in the next 5 to 10 seconds. If the predicted deviation is ≥10mm, start pre-adjustment in advance to avoid further deviation.

[0078] Step 6: Fault Alarm and Data Traceability

[0079] S6.1 Alarm Trigger: When the third-level deviation lasts for ≥3 seconds, the deviation adjustment mechanism fails (abnormal motor current), or the sensor fails, an audible and visual alarm is triggered and pushed to the central control system;

[0080] S6.2 Data Traceability: Records all deviation events, control processes, and operating parameters, with a retention period of ≥1 year, supporting fault tracing and process optimization.

[0081] In one embodiment, the method is applicable to coal and coke conveyor belts with a bandwidth of 500-2000mm, a belt speed of 1-3m / s, and a rated load of 0-500t / h, with a deviation control accuracy of ±5mm and a control response time of ≤0.5s.

[0082] In one embodiment, the detection frequency of the laser displacement sensor array in step S1.1 is 100Hz, the sampling frame rate of the industrial camera is 30fps, and the offset detection accuracy after data fusion is ≤±1mm.

[0083] In one embodiment, the training samples of the machine learning model in step S2.3 are ≥100,000 sets, covering all working conditions of load 0~100%, speed 1~3m / s, and offset 0~100mm, and the accuracy of the model output adjustment parameters is ≥98%.

[0084] In one embodiment, the adjustment of the alignment roller in step S3.1 is driven by a servo motor with an angle adjustment accuracy of ±0.1°, and the displacement of the alignment roller is driven by a ball screw with a displacement accuracy of ±0.1mm.

[0085] In one embodiment, in step S4.1, the adaptive dust removal vent uses an electric actuator to adjust the angle, with an adjustment range of 0~30° and an adjustment accuracy of ±0.5°. The air volume is controlled by a variable frequency fan, with a control range of 500~2000m³ / h.

[0086] In one embodiment, the iteration cycle of self-learning optimization in step S5.1 is 24 hours, and the model adjustment success rate is improved by ≥0.5% after each iteration, which is suitable for the working condition of belt aging (elastic modulus decay of 0 to 20%).

[0087] In one embodiment, the trend prediction in step S5.2 adopts an LSTM neural network model, and the input parameters include real-time offset, offset rate, load, speed, and tension force, with a prediction accuracy of ≥95%.

[0088] In one embodiment, the alarm information in step S6.1 includes the deviation level, the location of occurrence, the current operating condition, and the suggested control scheme, and the response time to push it to the central control system is ≤1s.

[0089] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various variations or substitutions within the technical scope disclosed in the present invention, and these should all be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. An intelligent anti-deviation adaptive control method for a coal and coke conveyor belt, characterized in that: Includes the following steps: Step 1: Multi-dimensional data collection (real-time sensing) S1.1 Misalignment Detection: One set of laser displacement sensor arrays (detection accuracy ±1mm) is arranged at the head, middle and tail of the belt conveyor. At the same time, an industrial camera is deployed for visual recognition to collect the belt lateral offset (0~100mm), offset direction (left / right offset) and offset rate (mm / s) in real time. S1.2 Operating Condition Data Acquisition: The real-time load of the belt (0~100% of the rated load) is collected by the load sensor, the belt running speed (1~3m / s) is collected by the speed sensor, the tilt angle of the idler roller is collected by the tilt sensor, and the ambient dust concentration (0~100mg / m³) is collected by the dust sensor. S1.3 Status Acquisition: Acquire status parameters such as the current angle, drive motor current, and belt tension (0~50kN) of the alignment mechanism (alignment roller, alignment drum); Step 2: Data Preprocessing and Intelligent Analysis S2.1 Data Cleaning: Remove sensor outliers (such as abrupt data changes caused by dust obstruction), and fuse and calibrate laser and vision inspection data to ensure that the offset detection error is ≤ ±1mm; S2.2 Deviation Level Determination: Deviation levels are determined based on the amount of offset. Level 1 (early warning): 5-20mm, initiate pre-regulation; Level 2 (Mild): 20-50mm, initiate normal control; Level 3 (Severe): 50mm, activate emergency control and alarm; S2.3 Intelligent Modeling and Analysis: Based on machine learning algorithms (random forest), a correlation model of "offset - load - speed - tension - adjustment parameters" is established, and the optimal adjustment strategy (adjustment roller angle, correction roller displacement, tension adjustment value) is output. Step 3: Adaptive Adjustment Execution S3.1 Tiered Regulation: Level 1 deviation: Adjust the tilt angle of the adjustment roller (adjustment range 0~15°, adjustment step 0.5°), and simultaneously adjust the angle of the adaptive dust removal air outlet to avoid uneven airflow. Secondary deviation: Based on the adjustment of the deviation idler, fine-tune the lateral displacement of the correction roller (adjustment range 0~50mm, adjustment step 1mm) to adapt to load / speed changes; Level 3 belt misalignment: immediately reduce belt speed to ≤1m / s, and simultaneously activate the double-sided adjustment rollers and correction drum for linkage adjustment, with adaptive tension compensation (adjustment range ±5kN). S3.2 Real-time feedback: The offset is collected every 0.1s during the adjustment process. If the offset does not decrease after adjustment, the adjustment parameters are automatically iterated and optimized until the offset is ≤5mm. Step 4: Coordinated Control of Dust Removal and Deviation Prevention S4.1 Adaptive Dust Removal: Adjust the angle (0~30°) and air volume (500~2000m³ / h) of the dust removal vent according to the direction / amount of belt offset, so that the airflow is perpendicular to the belt surface and avoids the airflow causing material to be unbalanced; S4.2 Sensor Dust Prevention: Adjust the dust removal air vents to perform directional blowing on the laser / vision sensor area, with a blowing air volume of 50-100 m³ / h, to ensure sensor cleanliness and detection accuracy ≥99%; Step 5: Self-learning and trend prediction S5.1 Self-learning optimization: Daily review of the day's deviation data, control effect, and operating parameters, iteratively update the machine learning model, and optimize the adjustment parameters under different operating conditions; S5.2 Trend Prediction: Based on historical data and real-time operating conditions, predict the belt deviation trend in the next 5 to 10 seconds. If the predicted deviation is ≥10mm, start pre-adjustment in advance to avoid further deviation. Step 6: Fault Alarm and Data Traceability S6.1 Alarm Trigger: When the third-level deviation lasts for ≥3 seconds, the deviation adjustment mechanism fails (abnormal motor current), or the sensor fails, an audible and visual alarm is triggered and pushed to the central control system; S6.2 Data Traceability: Records all deviation events, control processes, and operating parameters, with a retention period of ≥1 year, supporting fault tracing and process optimization.

2. The intelligent anti-deviation adaptive control method for a coal and coke conveyor belt according to claim 1, characterized in that: The method is applicable to coal and coke conveyor belts with a bandwidth of 500-2000mm, a belt speed of 1-3m / s, and a rated load of 0-500t / h. The deviation control accuracy is ±5mm, and the control response time is ≤0.5s.

3. The intelligent anti-deviation adaptive control method for a coal and coke conveyor belt according to claim 1, characterized in that: In step S1.1, the detection frequency of the laser displacement sensor array is 100Hz, the sampling frame rate of the industrial camera is 30fps, and the offset detection accuracy after data fusion is ≤±1mm.

4. The intelligent anti-deviation adaptive control method for a coal and coke conveyor belt according to claim 1, characterized in that: In step S2.3, the training samples of the machine learning model are ≥100,000 sets, covering all working conditions of load 0-100%, speed 1-3m / s, and offset 0-100mm, and the accuracy of the model output adjustment parameters is ≥98%.

5. The intelligent anti-deviation adaptive control method for a coal and coke conveyor belt according to claim 1, characterized in that: In step S3.1, the adjustment of the alignment roller is driven by a servo motor with an angle adjustment accuracy of ±0.1°, and the displacement of the alignment roller is driven by a ball screw with a displacement accuracy of ±0.1mm.

6. The intelligent anti-deviation adaptive control method for a coal and coke conveyor belt according to claim 1, characterized in that: In step S4.1, the adaptive dust removal vent uses an electric actuator to adjust the angle, with an adjustment range of 0~30° and an adjustment accuracy of ±0.5°. The air volume is controlled by a variable frequency fan, with a control range of 500~2000m³ / h.

7. The intelligent anti-deviation adaptive control method for a coal and coke conveyor belt according to claim 1, characterized in that: The iteration cycle of self-learning optimization in step S5.1 is 24 hours. After each iteration, the model adjustment success rate increases by ≥0.5%, which is suitable for the working condition of belt aging (elastic modulus decay of 0-20%).

8. The intelligent anti-deviation adaptive control method for a coal and coke conveyor belt according to claim 1, characterized in that: In step S5.2, the trend prediction adopts an LSTM neural network model, and the input parameters include real-time offset, offset rate, load, speed, and tension force, with a prediction accuracy of ≥95%.

9. The intelligent anti-deviation adaptive control method for a coal and coke conveyor belt according to claim 1, characterized in that: The alarm information in step S6.1 includes the deviation level, the location of occurrence, the current operating condition, and the suggested control scheme. The response time to push the information to the central control system is ≤1 second.

Images