Tension optimization control system and method for tensioning device of belt conveyor in underground coal mine

Abstract

The present invention discloses a tension optimization control system and method for a tensioning device of a coal mine underground belt conveyor, which relates to the technical field of tensioning device optimization control; the method comprises: S1, real-time collection of multi-source tensioning data, and data preprocessing of the multi-source tensioning data; S2, predicting the coal flow load based on historical multi-source tensioning data, and taking dynamic compensation measures for the tensioning device; S3, combining the multi-source tensioning data with the coal flow load prediction results, collaboratively allocating the hybrid actuators, and automatically regulating based on the collaborative allocation results; S4, real-time detection of anomalies based on the multi-source tensioning data, and early warning; S5, performing safety risk assessment based on the real-time multi-source tensioning data, and taking adaptive control measures based on the safety risk assessment results and the anomaly detection results; the problem of failure caused by high-frequency pulsation and short-term impact of belt tension that is difficult to adjust in time is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of tensioning device optimization control, in particular to a tension optimization control system and method for a tensioning device of a belt conveyor in an underground coal mine. Background Art

[0002] Underground coal mine belt conveyors, essential equipment for continuous coal transportation, are a critical link connecting mining, transportation, and surface production. With the continuous advancement of automation and intelligentization in coal mines, conveying systems face ever-higher operational efficiency and safety requirements. Under complex and variable underground operating conditions, factors such as the coal flow load, drive start-up and shutdown, and ambient temperature fluctuations can cause frequent fluctuations in belt tension, resulting in high-frequency pulsation and short-duration shocks, which can easily cause belt failures such as skipping, slipping, and even breakage.

[0003] For example, the invention patent with publication number CN106325149B discloses a method for intelligent control data acquisition and processing of a belt conveyor during earth pressure balance shield tunneling. The method is based on an acquisition and processing system, which includes a signal processing unit, a power module, a display unit and interface unit, an A / D conversion module, and a CPU control module. The CPU control module includes a load mode judgment unit, a data optimization unit, and a data correction unit. The load mode judgment unit initializes the tension data C of the belt conveyor in the no-load state and determines its operating mode based on the load tonnage. The data optimization unit obtains the optimal tension value D based on the sensor data. The data correction unit compares the tension data C with the optimal tension value D and performs PID adjustment processing. The present invention collects data synchronously through grouping tension sensors, utilizes fuzzy control algorithms and PID adjustment for data processing, and adjusts the speed of the belt conveyor according to the real-time working load. The control is precise, convenient for operation at the construction site, and achieves the purpose of energy saving and consumption reduction.

[0004] For example, the invention patent with the publication number CN104991532B discloses a centralized control automation device for an underground transportation system, including a track-aligned belt, a first coal feeder anti-collapse device, a second coal feeder anti-blocking and anti-collapse device, a control terminal and a camera, wherein the control terminal is used to control the track-aligned belt, the first coal feeder anti-collapse device and the second coal feeder anti-blocking and anti-collapse device. The present invention realizes remote monitoring and control of the transportation system and the collection and processing of relevant audio and video data by combining the traditional underground transportation system with a modern control system and a computer system. The present invention has high automation efficiency and reliable control results. By installing cameras at important locations on each belt and coal feeder, the image of each video is centrally transmitted to the display, which facilitates centralized observation by personnel, and the coal feeder is turned on according to the coal storage situation in the coal bin to reduce the occurrence of bin blockage.

[0005] However, in the process of implementing the technical solutions of the invention in the embodiments of the present application, the present application found that the above technology has at least the following technical problems:

[0006] Factors such as coal flow, load, and drive start and stop will cause high-frequency pulsation and short-term impacts in the belt tension. The underground environment is changeable and highly disturbed. The existing control system responds slowly and is difficult to adjust in time, which can easily cause belt jumping, slipping, and even breakage, affecting transportation safety and efficiency.

[0007] Therefore, in response to the above problems, there is an urgent need for a tension optimization control system and method for the tensioning device of a coal mine underground belt conveyor. Summary of the Invention

[0008] Technical problems solved

[0009] In response to the deficiencies in the prior art, the present invention provides a tension optimization control system and method for the tensioning device of a coal mine underground belt conveyor, which solves the problem of failures caused by high-frequency pulsation and short-term impact of the belt tension, which is difficult to adjust in time.

[0010] Technical Solution

[0011] To achieve the above objectives, the present invention is implemented through the following technical solutions: a method for optimizing the tension control of a tensioning device of a coal mine underground belt conveyor, comprising: S1, real-time collection of multi-source tensioning data, data preprocessing of the multi-source tensioning data, and establishment of a multi-source tensioning database; S2, predicting the coal flow load based on historical multi-source tensioning data, and taking dynamic compensation measures for the tensioning device according to the coal flow load prediction results; S3, combining the multi-source tensioning data with the coal flow load prediction results, collaboratively allocating the hybrid actuators, and automatically regulating based on the collaborative allocation results; S4, real-time detection of anomalies based on multi-source tensioning data, and early warning based on the anomaly detection results; S5, performing safety risk assessment based on real-time multi-source tensioning data, and taking adaptive control measures based on the safety risk assessment results and the anomaly detection results.

[0012] Furthermore, the specific process of real-time collection of multi-source tensioning data, data preprocessing of the multi-source tensioning data, and establishment of a multi-source tensioning database is as follows: real-time collection of multi-source tensioning data, including: measuring the tension of the belt using an FBG optical fiber strain sensor; collecting tension data and vibration data using a vector tension scale and a MEMS three-axis vibration sensor; collecting coal flow load data using a coal flow sensor and a visual monitoring camera, including: instantaneous coal flow rate, coal flow pulsation spectrum, and visual coal flow image; collecting current and torque data using a current sensor and a torque sensor; collecting the speed data of the moving wheel using a speed sensor; suppressing high-frequency noise through wavelet filtering, eliminating sudden over-range signals and unreasonable extreme multi-source tensioning data, filling in occasional missing data points using the Kalman filter interpolation method, and standardizing and normalizing the multi-source tensioning data; synchronously transmitting the multi-source tensioning data to the edge computing platform at the microsecond level and storing it in the multi-source tensioning database.

[0013] Furthermore, the specific process of predicting coal flow load based on historical multi-source tension data is as follows: obtain current coal flow load data, process the visual coal flow image through image processing algorithm, extract the mean of ROI area to obtain the characteristic mean of current coal flow monitoring image; based on the length of sliding time window, calculate the first-order derivative of the characteristic mean of current coal flow monitoring image with respect to time to obtain the characteristic mean change rate of coal flow monitoring image; multiply the current coal flow load data by the time series weight factor to obtain the time series weighted value, and based on the length of sliding time window, accumulate the time series weighted values ​​of all sampling points and divide them by the length of sliding time window to obtain the time series weighted sliding mean; multiply the characteristic mean change rate of coal flow monitoring image by the visual mutation response weight factor to obtain the visual mutation response term; add the time series weighted sliding mean and the visual mutation response term to obtain the coal flow load prediction value.

[0014] Furthermore, the specific process of taking dynamic compensation measures for the tensioning device according to the coal flow load prediction results is as follows: according to the coal flow load prediction value, identify the sudden increase and decrease of the coal flow, and adjust the movement amplitude and speed of the tensioning cylinder and servo screw in advance; if it is predicted that the coal flow increase value exceeds the sudden increase threshold within the next 200 milliseconds, automatically increase the tensioning force setting value, and increase the drive motor torque and power in advance; when the coal flow load prediction value suddenly increases and the tension and vibration data exceed the tension and vibration average values ​​respectively, issue emergency tensioning and speed limit instructions in advance, and coordinate safety braking; if the coal flow decrease value is predicted to exceed the sudden decrease threshold, automatically reduce the tension force; and when the real-time tension is lower than the safety threshold, issue a belt slip and belt break warning in advance, and automatically link the alarm and braking system; the coal flow load prediction value and the actual coal flow load data curve and abnormal points are automatically uploaded to the cloud platform to generate inspection suggestions and spare parts warnings.

[0015] Furthermore, combining the multi-source tensioning data with the coal flow load prediction results, the specific process of collaborative allocation of the hybrid actuator is as follows: select the electro-hydraulic servo cylinder as the coarse adjustment mechanism and the servo ball screw as the fine adjustment mechanism to construct a hybrid actuator; obtain the coal flow load prediction value, and convert the expected tension value into the target tension value through the regression fitting algorithm; subtract the current actual tension collected in real time from the target tension value to obtain the current tension compensation demand value; based on the sliding time window length, perform average filtering on the current tension compensation demand value, extract the smooth and slowly changing low-frequency component to obtain the low-frequency tension demand value, and multiply the low-frequency tension demand value with the electro-hydraulic servo cylinder gain weight factor to obtain the collaborative allocation value of the electro-hydraulic servo cylinder; subtract the low-frequency tension demand value from the current tension compensation demand value to obtain the high-frequency tension demand value, and multiply the high-frequency tension demand value with the servo ball screw gain weight factor to obtain the collaborative allocation value of the servo ball screw.

[0016] Furthermore, the specific process of automatic control based on the collaborative allocation result is as follows: based on the collaborative allocation value of the electro-hydraulic servo cylinder and the servo ball screw, the low-frequency and stable tensioning adjustment instructions are sent to the electro-hydraulic servo cylinder in real time through the edge controller; the high-frequency and fast compensation instructions are sent to the servo ball screw; the feedback signals of the electro-hydraulic servo cylinder and the servo ball screw are collected, and the feedback signals include actual thrust, displacement and response speed, and are compared with the corresponding collaborative allocation values ​​respectively, and the gain weight factors of the electro-hydraulic servo cylinder and the servo ball screw are dynamically adjusted according to the comparison error; if slow execution, insufficient response and overshoot occur, the collaborative allocation value is automatically recalculated; when the actual tensioning effect after allocation is found to be abnormal, the speed limit and shutdown operations are immediately executed and an alarm is issued; all allocation instructions and feedback signals are archived in real time, and comprehensive analysis is performed to generate energy consumption analysis, life prediction and maintenance recommendations.

[0017] Furthermore, the specific process of real-time anomaly detection based on multi-source tensioning data is as follows: obtain the current actual tension of the belt and the belt vibration data at the current moment, calculate the average value and standard deviation of the tension in the recent period based on the sliding time window, and obtain the historical mean value of the tension and the historical standard deviation of the tension respectively; and calculate the average value and standard deviation of the vibration data in the recent period, and obtain the historical mean value of the vibration and the historical standard deviation of the vibration respectively; subtract the historical mean value of the tension from the current actual tension and take the absolute value to obtain the tension deviation value, divide the tension deviation value by the historical standard deviation of the tension to obtain the tension deviation degree, and multiply the tension deviation degree by the tension anomaly weight factor to obtain the tension anomaly value; subtract the historical mean value of the vibration from the belt vibration data at the current moment and take the absolute value to obtain the vibration deviation value, divide the vibration deviation value by the historical standard deviation of the vibration to obtain the vibration deviation degree, and multiply the vibration deviation degree by the vibration anomaly weight factor to obtain the vibration anomaly value; add the tension anomaly value and the vibration anomaly value to obtain the anomaly monitoring value.

[0018] Furthermore, the specific process of early warning based on abnormal detection results is as follows: when the abnormal monitoring value is less than or equal to the abnormal threshold, it is judged to be in normal state, routine monitoring is maintained, and multi-source tensioning data, abnormal monitoring values ​​and device status are regularly archived; if it is continuously judged to be without abnormality for a period of time, the threshold optimization operation is performed according to the actual monitoring situation; when the abnormal monitoring value is greater than the abnormal threshold, it is judged to be in abnormal state, abnormal warning is triggered, the sound and light alarm is started, and the alarm information, including the abnormal monitoring value, timestamp, tension and vibration data, is pushed to the operation and maintenance terminal; the emergency adjustment mode of the tensioning mechanism is started to quickly increase and relieve the tension; the deceleration and shutdown are automatically performed; and the abnormal threshold, tension abnormality weight factor and vibration abnormality weight factor are automatically adjusted according to the abnormal monitoring value and the actual abnormal feedback information.

[0019] Furthermore, a safety risk assessment is performed based on real-time multi-source tensioning data. The specific process of taking adaptive control measures is as follows: obtaining the speed data of the driving wheel and the driven wheel, calculating the ratio of the speed difference between the driving wheel and the driven wheel to the speed of the main wheel to obtain the current slip rate, and screening the maximum slip rate within the sliding time window length as the slip rate limit safety value; obtaining the current actual tension, and screening the maximum tension within the sliding time window length as the tension limit safety value; obtaining the belt vibration data at the current moment, and screening the maximum vibration data within the sliding time window length as the vibration limit safety value; adding the ratio of the current slip rate to the slip rate limit safety value, the ratio of the current actual tension to the tension limit safety value, and the ratio of the belt vibration data at the current moment to the vibration limit safety value, and then Divide by the constant three to obtain the comprehensive proportion average value as the comprehensive safety risk assessment value; continuously obtain the safety risk assessment value, if the comprehensive safety risk assessment value does not exceed the risk threshold for two consecutive times, maintain routine monitoring and multi-source tensioning data collection, and update the multi-source tensioning data and comprehensive safety risk assessment value in real time. If the comprehensive safety risk assessment value does not exceed the risk threshold for four consecutive times, adaptively fine-tune the risk threshold; if the comprehensive safety risk assessment value exceeds the risk threshold for two consecutive times, an audible and visual alarm reminder will be issued, and a braking command will be automatically issued to control the driving wheel and the driven wheel to brake synchronously; and the hybrid actuator will be controlled to perform rapid actions; the alarm and action log will be pushed to the terminal in real time; based on the abnormal monitoring value and the comprehensive safety risk assessment value, the abnormality will be automatically pushed to the operation and maintenance personnel via SMS, and the operation and maintenance suggestions and abnormal risk reports will be automatically generated.

[0020] The second aspect of the present invention provides a tension optimization control system for a tensioning device of a belt conveyor in an underground coal mine, comprising: a high-speed perception and real-time fusion module, a prediction and compensation control module, a hybrid actuator module, an abnormality adaptive module, and a safety protection and cloud collaborative operation and maintenance module: wherein the high-speed perception and real-time fusion module is used to collect multi-source tensioning data in real time, perform data preprocessing on the multi-source tensioning data, and establish a multi-source tensioning database; the prediction and compensation control module is used to predict the coal flow load based on historical multi-source tensioning data, and take dynamic compensation measures for the tensioning device according to the coal flow load prediction results; the hybrid actuator module is used to combine the multi-source tensioning data with the coal flow load prediction results, coordinately allocate the hybrid actuators, and perform automatic regulation based on the coordinated allocation results; the abnormality adaptive module is used to detect abnormalities in real time based on the multi-source tensioning data, and perform early warning according to the abnormality detection results; the safety protection and cloud collaborative operation and maintenance module is used to perform safety risk assessment based on real-time multi-source tensioning data, and take adaptive regulation measures based on the safety risk assessment results and the abnormality detection results.

[0021] Beneficial effects

[0022] The present invention has the following beneficial effects:

[0023] (1) The present invention achieves high-reliability, low-noise, microsecond-level synchronous data acquisition by integrating multiple types of sensors and multiple data filtering methods, greatly improving the accuracy and sensitivity of tension state monitoring.

[0024] (2) The present invention combines coal flow load prediction with an adaptive regulation mechanism to achieve early response and precise regulation of sudden working conditions, effectively prevent risks such as slippage and belt breakage, and ensure safe operation of the system.

[0025] (3) The present invention realizes large-scale rapid adjustment and high-precision compensation by adopting a hybrid execution of an electro-hydraulic servo cylinder and a servo ball screw, thereby improving the adjustment efficiency while optimizing energy consumption and equipment life.

[0026] (4) The present invention realizes real-time early warning and automatic protection by constructing anomaly detection and risk assessment based on multiple indicators, forms operation and maintenance suggestions and self-learning optimization, and comprehensively improves the intelligence and security of the system.

[0027] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 This is a flow chart of the tension optimization control method for the tensioning device of the belt conveyor in the coal mine;

[0029] Figure 2 This is the structure diagram of the tension optimization control system for the tensioning device of the belt conveyor in the coal mine;

[0030] Figure 3 It is a trend chart of abnormal monitoring values. DETAILED DESCRIPTION

[0031] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. As those skilled in the art will understand, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0032] See also Figure 1-Figure 3 , an embodiment of the present invention provides a technical solution: a tension optimization control system and method for a tensioning device of a coal mine underground belt conveyor, including: S1, real-time collection of multi-source tensioning data, data preprocessing of the multi-source tensioning data, and establishment of a multi-source tensioning database; S2, prediction of coal flow load based on historical multi-source tensioning data, and taking dynamic compensation measures for the tensioning device according to the coal flow load prediction results; S3, combining the multi-source tensioning data with the coal flow load prediction results, collaboratively allocating the hybrid actuators, and automatically regulating based on the collaborative allocation results; S4, real-time detection of anomalies based on multi-source tensioning data, and early warning according to the anomaly detection results; S5, safety risk assessment based on real-time multi-source tensioning data, and taking adaptive control measures based on the safety risk assessment results and the anomaly detection results.

[0033] Specifically, the specific process of real-time collection of multi-source tensioning data, data preprocessing of the multi-source tensioning data, and establishment of a multi-source tensioning database is as follows: real-time collection of multi-source tensioning data, including: measuring the tension of the belt using an FBG optical fiber strain sensor; collecting tension data and vibration data using a vector tension scale and a MEMS three-axis vibration sensor; collecting coal flow load data using a coal flow sensor and a visual monitoring camera, including: instantaneous coal flow rate, coal flow pulsation spectrum, and visual coal flow image; collecting current and torque data using a current sensor and a torque sensor; and collecting the speed data of the moving wheel using a speed sensor. When preprocessing the multi-source tensioning data, first, high-frequency noise is suppressed by wavelet filtering, and sudden over-range signals and unreasonable extreme multi-source tensioning data are eliminated. Subsequently, the Kalman filter interpolation method is used to fill in occasional missing data points, and the multi-source tensioning data is standardized and normalized. The multi-source tensioning data is synchronously transmitted to the edge computing platform at the microsecond level and stored in the multi-source tensioning database.

[0034] In this implementation plan, multi-source belt tensioning data is collected in real time through multiple types of high-precision sensors. After filtering, denoising, anomaly removal, missing interpolation and standardization, the data is synchronously transmitted at the microsecond level and stored in the database. This not only significantly improves the accuracy and completeness of the data, but also provides a high-quality, comprehensive data foundation for subsequent intelligent analysis, fault prediction and optimized control, supporting the system to achieve a higher level of intelligent and safe operation and maintenance.

[0035] Specifically, the specific process of predicting coal flow load based on historical multi-source tension data is as follows: obtain the current coal flow load data, and process the visual coal flow image through the image processing algorithm. The image processing algorithm is specifically as follows: grayscale, ROI region segmentation and ROI region grayscale mean calculation, extract the mean of the ROI region to obtain the characteristic mean of the current coal flow monitoring image; based on the length of the sliding time window, calculate the first-order derivative of the characteristic mean of the current coal flow monitoring image with respect to time to obtain the characteristic mean change rate of the coal flow monitoring image, which reflects the mutation characteristics of the coal flow distribution and flow state; the current The coal flow load data is multiplied by the time series weight factor to obtain the time series weighted value. Based on the length of the sliding time window, the time series weighted values ​​of all sampling points are accumulated and then divided by the length of the sliding time window to obtain the time series weighted sliding mean, thereby smoothing the coal flow changes and highlighting the trend characteristics; the characteristic mean change rate of the coal flow monitoring image is multiplied by the visual mutation response weight factor to obtain the visual mutation response term, which further enhances the system's perception of abnormal fluctuations in coal flow; taking into account the time series and visual changes, the time series weighted sliding mean and the visual mutation response term are added to obtain the coal flow load prediction value.

[0036] The specific formula for the coal flow load prediction value is:

[0037] ;

[0038] Where, Indicates the future The coal flow load prediction value at the moment is used to predict the coal flow load for subsequent tensioning and compensation control, and respond to possible fluctuations in advance; Indicates the length of the sliding time window; Indicates the current coal flow load data; It represents the feature mean of the current coal flow monitoring image, which serves as a leading signal for judging the sudden change of coal flow state; The rate of change of the mean value of the coal flow monitoring image features indicates how quickly the image features change over time. Rapid changes mean that the coal flow increases or decreases suddenly, which has an immediate impact on the load. It represents the time series weight factor. The sliding window prediction is made by combining the historical coal flow load data with the actual future coal flow load data. The genetic algorithm is used to adjust the time series weight factor that minimizes the sliding window prediction error. The optimal time series weight factor is in the range of 0.2 to 0.6. It represents the visual mutation response weight factor, which is obtained by least squares fitting based on the change rate data of the feature mean of historical coal flow monitoring images, and its value range is between 0.5 and 2.0; It represents the time series weighted sliding mean, reflecting the short-term change trend of the coal flow load measured by the coal flow sensor, and is the basic prediction of the upcoming load; It represents the visual mutation response term. By quantifying the first-order temporal rate of change of the coal flow monitoring image features, it captures the load jump caused by coal flow mutation and improves the ability to respond in advance to sudden coal flow conditions.

[0039] This implementation integrates real-time coal flow load, image feature changes, and weight distribution, and utilizes a sliding weighted mean combined with visual mutation response to achieve dynamic and accurate prediction of coal flow load. This approach not only improves the system's ability to perceive and respond to coal flow fluctuations, but also enables early identification of sudden coal flow anomalies, enabling timely activation of downstream tensioning control and protection mechanisms. This provides a solid data and algorithmic foundation for proactive adjustment and intelligent linkage of underground conveying systems.

[0040] Specifically, the specific process of taking dynamic compensation measures for the tensioning device according to the coal flow load prediction results is as follows: according to the coal flow load prediction value, identify the sudden increase and decrease of the coal flow, adjust the action amplitude and speed of the tensioning cylinder and servo screw in advance, and realize accurate compensation for different disturbances; if it is predicted that the increase value of the coal flow in the next 200 milliseconds exceeds the sudden increase threshold, the tensioning force setting value is automatically increased, and the torque and power of the driving motor are increased in advance, thereby effectively preventing the risk of belt loosening and slipping caused by the sudden increase of coal flow; when the coal flow load prediction value increases suddenly and the tension and vibration data are respectively When the tension and vibration exceed the average value, emergency tensioning and speed limit instructions are issued in advance, and safety braking is coordinated to ensure the safe operation of the conveying system to the maximum extent; if the predicted coal flow reduction value exceeds the sudden reduction threshold, the tension force is automatically reduced to avoid belt damage caused by excessive tension; and when the real-time tension is lower than the safety threshold, belt slip and belt break warnings are issued in advance, and the alarm and braking systems are automatically linked; the coal flow load prediction value and the actual coal flow load data curve and abnormal points are automatically uploaded to the cloud platform to generate inspection suggestions and spare parts warnings, further improving the intelligence and initiative of operation and maintenance.

[0041] In this implementation plan, by predicting the coal flow load in advance, identifying mutation trends, and dynamically adjusting the tensioning and driving parameters, rapid response and protection to abnormal working conditions are achieved. This method can not only effectively prevent major faults such as belt jumping, slipping, and breakage, but also support real-time uploading of key data and early warning information to the cloud, realizing the intelligent push of abnormal inspection suggestions and spare parts early warnings, and significantly improving the system's inherent safety, remote management capabilities, and intelligent operation and maintenance level.

[0042] Specifically, combining multi-source tensioning data with coal flow load prediction results, the specific process of collaborative allocation of hybrid actuators is as follows: select the electro-hydraulic servo cylinder as the coarse adjustment mechanism, responsible for large-amplitude, low-frequency tension adjustment; select the servo ball screw as the fine adjustment mechanism, specifically for small-amplitude, high-frequency rapid tension compensation, to build a hybrid actuator; obtain the coal flow load prediction value, convert the expected tension value into the target tension value through the regression fitting algorithm, and thus determine the target tension; use the target tension value to subtract the current actual tension collected in real time to obtain the current tension compensation demand value, which is used for subsequent adjustment Provide a quantitative basis; based on the length of the sliding time window, perform average filtering on the current tension compensation demand value, extract the smooth and slowly changing low-frequency component to obtain the low-frequency tension demand value, and multiply the low-frequency tension demand value with the electro-hydraulic servo cylinder gain weight factor to obtain the coordinated distribution value of the electro-hydraulic servo cylinder, thereby ensuring the stability and efficiency of the main adjustment direction; subtract the low-frequency tension demand value from the current tension compensation demand value to obtain the high-frequency tension demand value, and multiply the high-frequency tension demand value with the servo ball screw gain weight factor to obtain the coordinated distribution value of the servo ball screw, thereby achieving accurate and rapid response to high-frequency disturbances.

[0043] The specific formula for the collaborative allocation value is:

[0044] ;

[0045] Where, Indicates the current tension compensation demand value, represents the error between the expected and actual value, and is the direct driving signal for all control actions; Indicates the future The target tension value at the moment reflects the expected belt tension compensation demand; Indicates the current actual tension and represents the real-time stress status; Indicates the low-frequency tension demand value and provides smooth control instructions for the electro-hydraulic servo cylinder of the coarse adjustment mechanism; Indicates the high-frequency tension demand value, providing smooth control instructions for the servo ball screw of the fine-tuning mechanism; Represents the electro-hydraulic servo cylinder gain weight factor, which is obtained by fitting the optimal gain between actual tension output and expected tension using a linear regression algorithm based on the historical belt tension data and the coordinated distribution value of the electro-hydraulic servo cylinder. The value range is between 0.5 and 2.0. The servo ball screw gain weight factor is based on the high-frequency tension demand value and the coordinated distribution value of the servo ball screw. It is obtained by taking the high-frequency response and fitting it with the least squares method. The value range is between 0.8 and 3.0. Indicates the coordinated distribution value of the electro-hydraulic servo cylinder, used for large-scale and slow-changing tension adjustment; Indicates the coordinated distribution value of the servo ball screw, which is used to compensate for small and fast changes in dynamics.

[0046] In this implementation plan, based on the coal flow load prediction and multi-source tensioning data, a regression algorithm is used to dynamically allocate the adjustment tasks of the coarse and fine adjustment mechanisms, realizing the coordinated control of the hybrid actuator and precise tension compensation, and improving the dynamic adaptability and overall control accuracy of the tensioning system; through the frequency division and decoupling strategy, coarse adjustment is responsible for low-frequency large-scale adjustment, and fine adjustment is responsible for high-frequency rapid response, which significantly improves the system's response speed, control accuracy and dynamic stability, and effectively alleviates the risks of belt jumping, slipping, etc. caused by tension fluctuations.

[0047] Specifically, the specific process of automatic control based on the collaborative allocation results is as follows: based on the collaborative allocation value of the electro-hydraulic servo cylinder and the servo ball screw, the low-frequency and stable tensioning adjustment instructions are sent to the electro-hydraulic servo cylinder in real time through the edge controller; the high-frequency and fast compensation instructions are sent to the servo ball screw to realize the division of labor and efficient collaboration of the coarse adjustment and fine adjustment mechanisms; the feedback signals of the electro-hydraulic servo cylinder and the servo ball screw are collected, and the feedback signals include actual thrust, displacement and response speed, and are compared with the corresponding collaborative allocation values ​​respectively. The gain weight factors of the electro-hydraulic servo cylinder and the servo ball screw are dynamically adjusted according to the comparison error, where the error is strictly limited to within 5% to ensure control accuracy and response sensitivity; if slow execution, insufficient response and overshoot occur, the collaborative allocation value is automatically recalculated; when the actual tensioning effect after allocation is found to be abnormal, the speed limit and shutdown operations are immediately executed and an alarm is issued to ensure equipment safety; all allocation instructions and feedback signals are archived in real time, and comprehensive analysis is performed to generate energy consumption analysis, life prediction and maintenance recommendations, providing data support for equipment health management and intelligent operation and maintenance decisions.

[0048] This implementation achieves real-time collaboration and dynamic optimization of the coarse and fine adjustment mechanisms through automated control and feedback correction of the coordinated allocation value. By acquiring feedback signals such as the actuator's real-time thrust, displacement, and response speed, the gain parameters are adaptively adjusted based on actual errors, ensuring tensioning adjustment accuracy and execution consistency. This mechanism not only enhances the actuator's responsiveness and stability under complex operating conditions, but also effectively avoids execution anomalies such as overshoot and hysteresis. It also provides valuable operational data support and decision-making basis for subsequent energy consumption analysis, fault identification, lifespan prediction, and intelligent operation and maintenance strategies.

[0049] Specifically, the real-time anomaly detection process based on multi-source tensioning data is as follows: The current actual belt tension and the current belt vibration data are obtained. Based on a sliding time window, the average and standard deviation of the tension over the recent period are calculated to obtain the historical tension mean and tension standard deviation, respectively. Furthermore, the average and standard deviation of the vibration data over the recent period are calculated to obtain the historical vibration mean and vibration standard deviation, respectively. This statistical feature extraction not only dynamically reflects the normal fluctuation range of tension and vibration, but also provides a reliable baseline for subsequent anomaly measurement. The tension deviation value is obtained by subtracting the historical tension mean from the current actual tension and taking its absolute value. The tension deviation value is divided by the historical tension standard deviation to obtain the tension deviation degree. The tension deviation degree is multiplied by the tension anomaly weight factor to obtain the tension anomaly value. The vibration deviation value is obtained by subtracting the historical vibration mean from the current belt vibration data and taking its absolute value. The vibration deviation degree is obtained by dividing the vibration deviation value by the historical vibration standard deviation to obtain the vibration deviation degree. The vibration anomaly value is multiplied by the vibration anomaly weight factor to obtain the vibration anomaly value. The tension anomaly value and the vibration anomaly value are added together to obtain the anomaly monitoring value. By integrating the two types of abnormal signals, abnormal operating conditions during operation can be sensitively identified, enabling real-time dynamic monitoring of equipment health status and early risk detection.

[0050] The specific formula for the abnormal monitoring value is:

[0051] ;

[0052] Where, Indicates the abnormal monitoring value, indicating the degree to which the current tension and vibration signals deviate from the normal working conditions. It is a quantitative value used by the system to determine whether there is an abnormality. It is used to detect short-term abnormalities and sudden trend changes, facilitating early warning. Indicates the current actual tension, used to detect tension fluctuations and anomalies; Indicates the historical average value of tension, reflecting the stable center value of recent belt tension, and is used to determine whether the current tension deviates from the normal value; Indicates the historical standard deviation of tension, which is used to measure the normal fluctuation range of tension. The larger the standard deviation, the more severe the tension fluctuation. Indicates the current belt vibration data, reflecting whether the belt has abnormal vibration, impact or failure; Indicates the historical mean value of vibration, which is used to determine whether the current vibration level deviates from the normal range; Indicates the historical standard deviation of vibration, which is used to measure the fluctuation amplitude of vibration under normal working conditions; The tension anomaly weight factor is obtained by fitting the historical tension data and the historical anomaly monitoring value dataset through the minimum error regression algorithm. It is used to adjust the influence of the tension term on the anomaly monitoring value and reflects the sensitivity and importance of tension changes to anomalies. The value range is between 0.3 and 0.8. Represents the vibration anomaly weight factor, which is obtained by fitting the minimum error regression algorithm based on historical vibration data and historical anomaly monitoring value datasets. It is used to adjust the impact of vibration items on anomaly monitoring values ​​and reflects the sensitivity and importance of vibration changes to anomalies. The value range is between 0.2 and 0.7.

[0053] The abnormal tension weight factor is set to 0.6, and the abnormal vibration weight factor is set to 0.4. When the weight factors are the same, the abnormal monitoring value at each moment is calculated under the condition that the actual belt tension and vibration data change over time. This is shown in Table 1.

[0054] Table 1 Abnormal monitoring value data table

[0055] Sampling point <![CDATA[Normalized T m (t)]]> Normalized a(t) <![CDATA[μ T ]]> <![CDATA[σ T ]]> <![CDATA[μ a ]]> <![CDATA[σ a ]]> <![CDATA[S a (t)]]> 1 0.65 0.55 0.6 0.07 0.5 0.08 0.679 2 0.78 0.62 0.6 0.07 0.5 0.08 2.143 3 0.58 0.47 0.6 0.07 0.5 0.08 0.321 4 0.72 0.67 0.6 0.07 0.5 0.08 1.879 5 0.60 0.80 0.6 0.07 0.5 0.08 1.500

[0056] like Figure 3 As shown in the figure, it is an abnormal monitoring value trend diagram provided by the embodiment of the present application. The horizontal axis in the figure is the sampling time point, the vertical axis is the abnormal monitoring value, and the abnormal warning threshold is set to 1. According to Table 1 and Figure 3 It can be seen that when the weight factors are the same, the changing trend of the abnormal monitoring values ​​is as time changes and the current actual tension and vibration data of the belt also changes.

[0057] In this implementation, by performing sliding window statistics and standardized deviation analysis on tension and vibration data, not only can the degree of anomaly be quantified in real time, but the historical baseline can also be adaptively updated and the threshold can be dynamically optimized, thereby achieving anomaly detection and early warning with higher sensitivity and low false alarm rate in multi-source data fusion scenarios; with the help of this mechanism, faults can be accurately captured in the budding stage and automatically linked to protective actions, greatly improving the early protection capabilities and overall operational reliability of the underground transportation system.

[0058] Specifically, the specific process of early warning based on abnormal detection results is as follows: when the abnormal monitoring value is less than or equal to the abnormal threshold, it is judged to be in normal state, routine monitoring is maintained, and multi-source tensioning data, abnormal monitoring values ​​and device status are regularly archived to provide historical data support for subsequent equipment health management; if it is continuously judged to be without abnormality for a period of time, the threshold optimization operation is performed according to the actual monitoring situation, so as to adaptively improve the sensitivity and robustness of the system; when the abnormal monitoring value is greater than the abnormal threshold, it is judged to be in abnormal state, triggering abnormal warning, starting the sound and light alarm, and pushing the alarm information, including abnormal monitoring value, timestamp, tension and vibration data to the operation and maintenance terminal to realize real-time notification and traceability of fault events; starting the emergency adjustment mode of the tensioning mechanism, quickly increasing and relieving the tension to suppress potential risks; automatically decelerating and stopping to prevent further expansion of the fault; and automatically adjusting the abnormal threshold, tension abnormality weight factor and vibration abnormality weight factor according to the abnormal monitoring value and actual abnormal feedback information, so as to continuously optimize the early warning strategy and enhance adaptive and intelligent protection capabilities.

[0059] In this implementation plan, the system status is dynamically judged based on the anomaly detection results, which can trigger early warning and emergency regulation in a timely manner. It supports alarm information push, automatic parameter adjustment and intelligent optimization of thresholds. At the same time, the complete event link is uploaded to the cloud in real time and automatically archived, providing a basis for subsequent fault tracing, operation and maintenance decision-making and model self-learning, further improving the system's fault protection and self-adaptation capabilities.

[0060] Specifically, a safety risk assessment is performed based on real-time multi-source tensioning data, and the specific process of taking adaptive control measures is combined with the safety risk assessment results and the abnormality detection results. The specific process is as follows: obtain the speed data of the driving wheel and the driven wheel, calculate the ratio of the speed difference between the driving wheel and the driven wheel to the speed of the main wheel to obtain the current slip rate, and filter the maximum slip rate within the sliding time window length as the slip rate limit safety value to dynamically reflect the maximum slip risk of the equipment during operation; obtain the current actual tension, and filter the maximum tension within the sliding time window length as the tension limit safety value to ensure that the tension over-limit risk is monitored in time; obtain the belt vibration data at the current moment, and filter the maximum vibration data within the sliding time window length as the vibration limit safety value, so as to comprehensively evaluate the dynamic operation state of the belt; add the ratio of the current slip rate to the slip rate limit safety value, the ratio of the current actual tension to the tension limit safety value, and the ratio of the belt vibration data at the current moment to the vibration limit safety value, and then divide by the constant three to obtain The average value of the comprehensive proportion is used as the comprehensive safety risk assessment value to achieve unified risk quantification under multiple indicators; the safety risk assessment value is continuously obtained. If the comprehensive safety risk assessment value does not exceed the risk threshold for two consecutive times, the routine monitoring and multi-source tensioning data collection are maintained, and the multi-source tensioning data and the comprehensive safety risk assessment value are updated in real time. If the comprehensive safety risk assessment value does not exceed the risk threshold for four consecutive times, the risk threshold is adaptively fine-tuned; if the comprehensive safety risk assessment value exceeds the risk threshold for two consecutive times, the system's adaptability to different working conditions is improved, and an audible and visual alarm reminder is issued, and a braking command is automatically issued to control the driving wheel and the driven wheel to brake synchronously; and the hybrid actuator is controlled to perform rapid actions to ensure rapid response and equipment safety in abnormal situations; alarms and action logs are pushed to the terminal in real time; based on abnormal monitoring values ​​and comprehensive safety risk assessment values, abnormalities are automatically pushed to operation and maintenance personnel via SMS, and operation and maintenance suggestions and abnormal risk reports are automatically generated to achieve intelligent operation and maintenance and scientific decision support.

[0061] The specific formula for the comprehensive safety risk assessment value is:

[0062] ;

[0063] Where, Indicates the comprehensive safety risk assessment value, reflecting the overall risk level of the current belt operation, and is used to control extreme safety and overall health; Indicates the current slip rate, which is a direct indicator of slippage, sticking, and loss of step; Indicates the slip rate limit safety value; Indicates the current actual tension, reflects the tension on the belt in real time, and is the main basis for judging overload and slack; Indicates the tension limit safety value; Indicates the current belt vibration data, reflecting belt jump, impact, and abnormal vibration problems, and is an important means of monitoring early faults; Indicates the vibration limit safety value.

[0064] In this implementation plan, the safety ratio of slip rate, tension and vibration is calculated in real time, the comprehensive risk is dynamically assessed, and the abnormal detection results are combined to achieve graded response and adaptive control. When the risk increases, braking, tensioning and cloud alarms are automatically linked. After the risk decreases, the thresholds and control parameters are adaptively optimized, and complete data, analysis and suggestions are pushed to the operation and maintenance platform. This not only achieves accurate early warning and rapid protection, but also provides decision support for long-term trend evaluation, intelligent maintenance scheduling and spare parts management, and comprehensively improves the inherent safety and intelligent operation and maintenance capabilities of the system.

[0065] Reference Figure 2 As shown, the second aspect of the present invention provides a tension optimization control system for a tensioning device of a coal mine underground belt conveyor, which is applied to the above-mentioned tension optimization control method for the tensioning device of a coal mine underground belt conveyor, including: a high-speed perception and real-time fusion module, a prediction compensation control module, a hybrid actuator module, an abnormality adaptive module and a safety protection and cloud collaborative operation and maintenance module: wherein the high-speed perception and real-time fusion module is used to collect multi-source tensioning data in real time, perform data preprocessing on the multi-source tensioning data, and establish a multi-source tensioning database; the prediction compensation control module is used to predict the coal flow load based on historical multi-source tensioning data, and take dynamic compensation measures for the tensioning device according to the coal flow load prediction results; the hybrid actuator module is used to combine the multi-source tensioning data with the coal flow load prediction results, coordinately allocate the hybrid actuators, and automatically control based on the coordinated allocation results; the abnormality adaptive module is used to detect abnormalities in real time based on the multi-source tensioning data, and provide early warning according to the abnormality detection results; the safety protection and cloud collaborative operation and maintenance module is used to perform safety risk assessment based on the real-time multi-source tensioning data, and take adaptive control measures based on the safety risk assessment results and the abnormality detection results.

[0066] In this implementation plan, multi-source data acquisition, intelligent prediction and compensation, hybrid actuator collaboration, real-time abnormal adaptation and cloud-based secure operation and maintenance are integrated, which can complete tensioning state perception, prediction, allocation, regulation and cloud-based closed-loop feedback at the millisecond level, realizing high-precision tension control and rapid risk response, and supporting remote fault diagnosis, energy consumption analysis and life prediction, thus comprehensively improving the intelligence level, operational reliability and inherent safety of the underground transportation system.

[0067] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0068] The preferred embodiments of the present invention disclosed above are intended only to help illustrate the present invention. The preferred embodiments do not describe all details in detail, nor do they limit the invention to the specific embodiments described. As those skilled in the art will appreciate, numerous modifications and variations are possible based on the contents of this specification. These embodiments are selected and described in detail in this specification to better explain the principles and practical applications of the present invention, thereby enabling those skilled in the art to better understand and utilize the present invention. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for optimizing the tension of a tensioning device of a belt conveyor in an underground coal mine, characterized in that: include: S1, real-time collection of multi-source tensioning data, data preprocessing of the multi-source tensioning data, and establishment of a multi-source tensioning database; S2, predicting the coal flow load based on historical multi-source tensioning data, and taking dynamic compensation measures for the tensioning device according to the coal flow load prediction results; The specific process of predicting coal flow load based on historical multi-source tension data is as follows: Obtain the current coal flow load data, process the visual coal flow image through the image processing algorithm, extract the mean value of the ROI area to obtain the feature mean of the current coal flow monitoring image; Based on the length of the sliding time window, the first-order derivative of the characteristic mean of the current coal flow monitoring image with respect to time is calculated to obtain the characteristic mean change rate of the coal flow monitoring image; The current coal flow load data is multiplied by the time series weight factor to obtain a time series weighted value. Based on the length of the sliding time window, the time series weighted values ​​of all sampling points are accumulated and then divided by the length of the sliding time window to obtain a time series weighted sliding mean. The characteristic mean change rate of the coal flow monitoring image is multiplied by the visual mutation response weight factor to obtain a visual mutation response term. The time series weighted sliding mean and the visual mutation response term are added to obtain the coal flow load prediction value. S3, combining multi-source tensioning data with coal flow load prediction results, coordinately allocates hybrid actuators, and performs automatic control based on the coordinated allocation results; The specific process of collaboratively allocating the hybrid actuators by combining the multi-source tensioning data with the coal flow load prediction results is as follows: An electro-hydraulic servo cylinder is selected as the coarse adjustment mechanism, and a servo ball screw is selected as the fine adjustment mechanism to construct a hybrid actuator; Obtain the coal flow load prediction value, and convert the expected tension value into the target tension value through the regression fitting algorithm; The current actual tension collected in real time is subtracted from the target tension value to obtain the current tension compensation requirement value; Based on the length of the sliding time window, the current tension compensation demand value is averaged and filtered to extract the smooth and slowly changing low-frequency components to obtain the low-frequency tension demand value. The low-frequency tension demand value is multiplied by the electro-hydraulic servo cylinder gain weight factor to obtain the coordinated distribution value of the electro-hydraulic servo cylinder. The high-frequency tension demand value is subtracted from the current tension compensation demand value to obtain the high-frequency tension demand value. The high-frequency tension demand value is multiplied by the servo ball screw gain weight factor to obtain the coordinated distribution value of the servo ball screw. S4, detects anomalies in real time based on multi-source tensioning data and issues early warnings based on the anomaly detection results; S5, conducts safety risk assessment based on real-time multi-source tensioning data, and takes adaptive control measures based on the safety risk assessment results and anomaly detection results.

2. The method for optimizing the tension of the tensioning device of a belt conveyor in an underground coal mine according to claim 1, characterized in that: The specific process of real-time acquisition of multi-source tensioning data, data preprocessing of the multi-source tensioning data, and establishment of a multi-source tensioning database is as follows: Real-time collection of multi-source tensioning data, including: measuring belt tension using FBG fiber strain sensors; collecting tension and vibration data using vector tension scales and MEMS triaxial vibration sensors; collecting coal flow load data using coal flow sensors and visual monitoring cameras, including: coal flow instantaneous flow rate, coal flow pulsation spectrum, and visual coal flow images; collecting current and torque data using current sensors and torque sensors; and collecting impeller speed data using speed sensors. Wavelet filtering is used to suppress high-frequency noise, eliminate sudden over-range signals and unreasonable extreme multi-source tension data, and use the Kalman filter interpolation method to fill in occasional missing data points, and standardize and normalize the multi-source tension data; The multi-source tensioning data is synchronously transmitted to the edge computing platform at the microsecond level and stored in the multi-source tensioning database.

3. The method for optimizing the tension of a tensioning device of a belt conveyor in an underground coal mine according to claim 1, characterized in that: The specific process of taking dynamic compensation measures for the tensioning device according to the coal flow load prediction result is as follows: According to the coal flow load prediction value, identify the sudden increase and decrease of coal flow, and adjust the movement amplitude and speed of the tensioning cylinder and servo screw in advance; If it is predicted that the coal flow increase value will exceed the sudden increase threshold within the next 200 milliseconds, the tension force setting value will be automatically increased, and the drive motor torque and power will be increased in advance; when the coal flow load prediction value suddenly increases and the tension and vibration data exceed the tension and vibration average values ​​respectively, emergency tension and speed limit instructions will be issued in advance, and safety braking will be coordinated; If the predicted coal flow reduction exceeds the sudden drop threshold, the tension force is automatically reduced; and when the real-time tension falls below the safety threshold, an early warning of belt slippage and breakage is issued, automatically linking the alarm and braking systems. The coal flow load prediction value, actual coal flow load data curve, and abnormal points are automatically uploaded to the cloud platform to generate inspection suggestions and spare parts warnings.

4. The method for optimizing the tension of a tensioning device of a belt conveyor in an underground coal mine according to claim 1, characterized in that: The specific process of automatic control based on the collaborative allocation results is as follows: Based on the coordinated distribution value of the electro-hydraulic servo cylinder and the servo ball screw, the edge controller sends low-frequency and stable tension adjustment instructions to the electro-hydraulic servo cylinder in real time; and sends high-frequency and fast compensation instructions to the servo ball screw. The feedback signals of the electro-hydraulic servo cylinder and servo ball screw, including actual thrust, displacement and response speed, are collected and compared with the corresponding coordinated distribution values. The gain weight factors of the electro-hydraulic servo cylinder and servo ball screw are dynamically adjusted according to the comparison error. If slow execution, insufficient response and overshoot occur, the coordinated distribution value is automatically recalculated. If the actual tensioning effect after distribution is found to be abnormal, speed limit and shutdown operations are immediately executed and an alarm is issued. All distribution instructions and feedback signals are archived in real time, and comprehensive analysis is performed to generate energy consumption analysis, life prediction and maintenance recommendations.

5. The method for optimizing the control of the tension of the tensioning device of a belt conveyor in an underground coal mine according to claim 1, characterized in that: The specific process of real-time anomaly detection based on multi-source tensioning data is as follows: Obtain the current actual belt tension and the current belt vibration data. Based on the sliding time window, calculate the average value and standard deviation of the tension in the recent period to obtain the historical mean value and historical standard deviation of the tension, respectively. Also calculate the average value and standard deviation of the vibration data in the recent period to obtain the historical mean value and historical standard deviation of the vibration, respectively. Subtract the historical mean tension value from the current actual tension and take the absolute value to obtain the tension deviation value, divide the tension deviation value by the historical standard deviation of the tension to obtain the tension deviation degree, and multiply the tension deviation degree by the tension anomaly weight factor to obtain the tension anomaly value; subtract the historical mean vibration value from the current belt vibration data and take the absolute value to obtain the vibration deviation value, divide the vibration deviation value by the historical standard deviation of the vibration to obtain the vibration deviation degree, and multiply the vibration deviation degree by the vibration anomaly weight factor to obtain the vibration anomaly value; add the tension anomaly value and the vibration anomaly value to obtain the anomaly monitoring value.

6. The method for optimizing the tension of a tensioning device of a belt conveyor in an underground coal mine according to claim 1, characterized in that: The specific process of early warning based on abnormal detection results is as follows: When the abnormal monitoring value is less than or equal to the abnormal threshold, it is determined to be normal, and routine monitoring is maintained. Multi-source tensioning data, abnormal monitoring values, and device status are regularly archived. If it is continuously determined to be normal for a period of time, the threshold optimization operation is performed based on the actual monitoring situation. When the abnormal monitoring value is greater than the abnormal threshold, it is judged as an abnormal state, triggering an abnormal warning, starting the sound and light alarm, and pushing the alarm information, including the abnormal monitoring value, timestamp, tension and vibration data, to the operation and maintenance terminal; starting the emergency adjustment mode of the tensioning mechanism to quickly increase and relieve the tension; automatically decelerating and stopping; and automatically adjusting the abnormal threshold, tension abnormality weight factor and vibration abnormality weight factor according to the abnormal monitoring value and actual abnormal feedback information.

7. The method for optimizing the control of the tension of the tensioning device of a belt conveyor in an underground coal mine according to claim 1, characterized in that: The specific process of performing safety risk assessment based on real-time multi-source tensioning data and taking adaptive control measures in combination with the safety risk assessment results and anomaly detection results is as follows: Obtain the speed data of the driving wheel and the driven wheel, calculate the ratio of the speed difference between the driving wheel and the driven wheel to the speed of the main wheel to obtain the current slip rate, and select the maximum slip rate within the sliding time window length as the slip rate limit safety value; Obtain the current actual tension and select the maximum tension within the sliding time window length as the tension limit safety value; obtain the belt vibration data at the current moment and select the maximum vibration data within the sliding time window length as the vibration limit safety value; The ratio of the current slip rate to the slip rate limit safety value, the ratio of the current actual tension to the tension limit safety value, and the ratio of the current belt vibration data to the vibration limit safety value are added together, and then divided by the constant three to obtain the comprehensive proportion average value as the comprehensive safety risk assessment value; Continuously obtain safety risk assessment values. If two consecutive comprehensive safety risk assessment values ​​do not exceed the risk threshold, maintain routine monitoring and multi-source tensioning data collection, and update the multi-source tensioning data and comprehensive safety risk assessment values ​​in real time. If four consecutive comprehensive safety risk assessment values ​​do not exceed the risk threshold, adaptively fine-tune the risk threshold. If the comprehensive safety risk assessment value exceeds the risk threshold for two consecutive times, an audible and visual alarm will be issued, and a braking command will be automatically issued to control the synchronous braking of the driving and driven wheels; And control the hybrid actuator to perform fast actions; push alarms and action logs to the terminal in real time; Based on the abnormal monitoring values ​​and comprehensive safety risk assessment values, abnormalities are automatically pushed to operation and maintenance personnel via SMS, and operation and maintenance suggestions and abnormal risk reports are automatically generated.

8. A system using the method for optimizing the control of the tension of a tensioning device of a coal mine underground belt conveyor according to any one of claims 1 to 7, comprising: High-speed perception and real-time fusion module, prediction compensation control module, hybrid actuator module, abnormal adaptation module and security protection and cloud collaborative operation and maintenance module, characterized by: The high-speed perception and real-time fusion module is used to collect multi-source tensioning data in real time, perform data preprocessing on the multi-source tensioning data, and establish a multi-source tensioning database; The prediction compensation control module is used to predict the coal flow load based on historical multi-source tensioning data and take dynamic compensation measures for the tensioning device according to the coal flow load prediction results; The hybrid actuator module is used to combine multi-source tensioning data with coal flow load prediction results to coordinately allocate hybrid actuators and perform automatic control based on the coordinated allocation results; The abnormality adaptive module is used to detect abnormalities in real time based on multi-source tensioning data and provide early warning according to the abnormality detection results; The security protection and cloud collaborative operation and maintenance module is used to perform security risk assessment based on real-time multi-source tension data, and take adaptive control measures based on the security risk assessment results and anomaly detection results.

Images