Integrated intelligent control system for belt conveyor of tunneling face

Abstract

The present application relates to the tunneling working face belt conveyor integrated intelligent control system, including the perception module, the edge calculation module and the execution module;The perception module is used for real-time acquisition of the cutting motor current of the tunneling machine, the cutting arm traction speed, and the belt speed and the coal amount distribution information on the belt conveyor;The edge calculation module internally deploys the collaborative optimization control algorithm based on the cutting-conveying coupling dynamic model, and the collaborative optimization control algorithm is used for executing the following process: the cutting motor current and the cutting arm traction speed collected by the perception module are used to estimate the coal rock cutting resistance coefficient online;According to the estimated coal rock cutting resistance coefficient, the cutting head geometric parameter and the rotational speed, the predicted coal flow in the future specified time domain is calculated;The predicted coal flow is used as the disturbance input;The present application can realize more intelligent comprehensive tunneling working face belt conveyor intelligent control.

Description

Technical Field

[0001] This invention relates to the field of control systems, and more specifically to an integrated intelligent control system for belt conveyors in tunneling faces. Background Technology

[0002] The tunneling face is the core operating area for coal mine roadway excavation. The coal cut by the tunneling machine is continuously transported out via belt conveyor, with the cutting and conveying processes closely linked. However, in existing technologies, the tunneling machine and belt conveyor are usually controlled independently, lacking an effective coordination mechanism. This leads to frequent changes in coal and rock conditions and drastic fluctuations in the cutting volume during the tunneling machine's cutting process. Meanwhile, belt conveyors often use constant speed or manual speed adjustment based on experience, making it impossible to adjust the conveying speed in a timely manner according to the actual coal flow rate. This can easily cause coal piling and spillage on the conveyor belt when the coal flow rate suddenly increases, and cause light load or idling when the coal flow rate decreases, affecting production safety and wasting a large amount of electrical energy.

[0003] Existing methods for detecting coal quantity mostly rely on lagging measurement methods such as belt scales. These methods can only determine the amount of coal that has already been transported downstream, and cannot predict future trends in coal drop. As a result, belt conveyor speed regulation is always in a passive response state, and dynamic balance between cutting and conveying capacity cannot be achieved.

[0004] In scenarios involving the joint operation of multi-stage conveying equipment, the speed decision of the upstream belt conveyor often fails to consider the real-time conveying capacity constraints of the downstream equipment, leading to frequent material blockages and overloads at transfer points, which affects the continuity and reliability of the entire transportation line.

[0005] Monitoring common faults such as conveyor belt slippage and misalignment usually relies on dedicated hardware such as slippage switches and misalignment sensors, which increases installation and maintenance costs and makes it difficult to distinguish between normal load fluctuations and torque abnormalities caused by faults, resulting in a high rate of false alarms and missed alarms.

[0006] While some model predictive control methods can comprehensively consider energy consumption and transmission capacity optimization, they rely on complex online optimization solutions, which require high computing power from edge computing modules and are difficult to operate stably in real time under the limited computing resources in underground coal mines. At the same time, the model parameters are mostly fixed values, which cannot adapt to the real-time changes in coal and rock hardness and looseness, and the prediction accuracy decreases significantly after long-term operation. To address the aforementioned issues, an integrated intelligent control system for belt conveyors in tunneling faces was proposed. Summary of the Invention

[0007] To address the shortcomings of existing technologies, this invention provides an integrated intelligent control system for belt conveyors in tunneling faces, comprising a sensing module, an edge computing module, and an execution module; The sensing module is used to collect real-time information on the cutting motor current and cutting arm traction speed of the tunneling machine, as well as the belt speed and coal distribution on the conveyor belt of the belt conveyor. The edge computing module deploys a collaborative optimization control algorithm based on a cut-and-transport coupled dynamic model. This algorithm is used to execute the following processes: The cutting motor current and cutting arm traction speed collected by the sensing module are used to estimate the coal and rock cutting resistance coefficient online. Based on the estimated coal and rock cutting resistance coefficient, cutting head geometric parameters and rotation speed, the expected coal flow rate in the future specified time domain is calculated; using the expected coal flow rate as a disturbance input, based on the preset dynamic energy consumption and wear model of the belt conveyor, a finite time domain optimization problem is constructed with the optimization objectives of conveying energy loss and conveyor belt wear rate, and the optimal conveyor belt speed reference trajectory is solved; The execution module generates speed control commands for the frequency converter of the belt conveyor based on the optimal conveyor belt speed reference trajectory, and outputs a traction speed suggestion value to the tunneling machine controller to achieve dynamic balance between coal cutting and conveying capacity.

[0008] Furthermore, the online estimation of the coal and rock cutting resistance coefficient is achieved through an extended Kalman filter based on the observation of the cutting motor load torque; The extended Kalman filter uses the coal and rock cutting resistance coefficient as a state variable and the cutting motor current and the cutting arm traction speed as measured values. Through state prediction and measurement update, it recursively calculates and obtains the estimated value of the coal and rock cutting resistance coefficient at the current moment.

[0009] Furthermore, the calculation process for the expected coal flow rate includes: Estimated coal and rock cutting resistance coefficient Cutting head radius R, cutting head rotation speed Based on the cutting depth d, the energy required for cutting a unit volume of rock is determined according to the cutting-to-energy ratio, and then the instantaneous coal flow rate is calculated by combining the real-time power of the cutting motor; the expected coal flow rate is then calculated. Obtained through the following formula: ; in, In order to cut the motor power, For reference, the uniaxial compressive strength of rock, The effective cutting area is related to the cutting head radius and cutting depth. A correction factor to account for the rock fracturing and loosening effect; The collaborative optimization control algorithm also maps the instantaneous volumetric flow rate to the expected coal drop flow rate sequence within the future time window based on the swing speed and cutting depth of the cutting arm.

[0010] Furthermore, the dynamic energy consumption and wear model of the belt conveyor will reduce the total power of the belt conveyor. Expressed as a function of belt speed v and material load per unit length q: ; in, to The identification coefficients are related to the structural parameters of the conveyor and the characteristics of the idler rollers, where g is the acceleration due to gravity and H is the material lifting height. The conveyor belt wear rate W(v,q) is quantitatively calculated based on the contact stress between the conveyor belt and the idler rollers and the relative slip velocity. Its expression is as follows: ; in, and The wear coefficient is related to the material properties of the conveyor belt and idler rollers; Total power Together with the wear rate W(v,q), it forms the comprehensive quantitative basis for energy loss and mechanical wear in the conveying process.

[0011] Furthermore, the finite-time domain optimization problem in the collaborative optimization control algorithm is implemented using a piecewise speed regulation strategy based on predicted load trends. The specific process is as follows: Based on the projected coal flow rate sequence, the estimated unit length load at each time point in the future prediction time domain is obtained by recursively calculating the projected coal flow rate according to the conservation of material mass. and take its maximum value. ; when If the load exceeds the critical load corresponding to the maximum allowable filling rate of the conveyor belt at the current belt speed, and the overload trend continues to rise, the control belt conveyor should be adjusted from the current belt speed. To the predetermined high belt speed Smooth transition; when When the load is below the lower limit of the efficient operating load range at the current belt speed, and the light load trend continues to decrease, control the belt conveyor to reduce the load from the current belt speed. To the predetermined low-speed economic gear Smooth transition; Maintain the current belt speed in all other cases; During the belt speed transition, acceleration and deceleration are constrained by a preset ramp function, while ensuring that the real-time filling rate on the conveyor belt does not exceed the maximum allowable value.

[0012] Furthermore, the edge computing module also receives material flow detection signals from the transfer machine or downstream belt conveyor, transforming the transport capacity boundary of the downstream belt conveyor into additional constraints for the segmented speed regulation strategy, so as to achieve speed coordination control between multi-level conveying equipment. The calculation process for transforming the transport capacity boundary of the downstream belt conveyor into constraint conditions is as follows: Based on the maximum permissible cross-sectional area of ​​the downstream belt conveyor Current real-time speed and material bulk density Calculate its maximum allowable transport flow rate. : ; When the estimated flow rate of the belt conveyor output to the transfer conveyor or downstream conveyor belt Exceed At that time, the segmented speed control strategy will output a speed reduction command to limit the speed. Not exceeding Output flow By predicting the load per unit length It is obtained by multiplying the speed of the conveyor belt.

[0013] Furthermore, the sensing module is also used to collect the current of the belt conveyor drive motor; the edge computing module also includes a conveyor belt slippage and misalignment monitoring unit, whose fault identification process is as follows: Based on the current signal and torque constant of the belt conveyor drive motor Calculate electromagnetic torque The calculation formula is: , where I is the effective value of the current calculated from the current signal; Based on the equivalent rotational inertia of the drive roller and the reducer The belt speed and drive drum radius of the belt conveyor Calculated roller angular velocity And the viscous friction coefficient B, using the driving drum dynamics equations to inversely calculate the actual load torque. The dynamic equation is: ; The estimated value of the actual load torque obtained by back-calculation is: ,in This refers to the angular acceleration of the drum. Based on the coal distribution information on the conveyor belt obtained by the sensing module, the mass of material per unit length is obtained. Combined with the mass per unit length of the conveyor belt Drive roller radius Conveyor inclination angle And the running resistance coefficient f, calculate the theoretical load torque caused by the weight of the material and the conveyor belt. The calculation formula is: ; Where g is the acceleration due to gravity; Calculate torque residual ; When torque residual If the absolute value exceeds the threshold calculated based on normal operating data and the duration exceeds the set confirmation time, it is determined that the conveyor belt has slipped or deviated.

[0014] Furthermore, the sensing module also includes a laser coal flow cross-section scanner; the edge computing module utilizes the measured coal flow volume retrieved by the laser scan. Compared with the expected coal flow rate Comparison of the loose effect correction factor Perform online calibration; The online calibration process is as follows: Collect a sequence of the ratios of the measured coal flow volume to the expected coal flow rate for the most recent L sampling periods. The arithmetic mean of the comparison value series is used to obtain the average ratio. : ; Use the current correction factor Updated to ; Where L is the sliding window length, the window length is selected according to the degree of coal flow fluctuation. When the fluctuation is severe, a smaller window is used for rapid response, and when the fluctuation is gentle, a larger window is used to smooth the noise.

[0015] Furthermore, the edge computing module includes a parameter self-learning unit, which utilizes real-time feedback of coal weighing data during system operation to assess the uniaxial compressive strength of the reference rock. and loose effect correction factor Perform online incremental updates; At the end of each weighing cycle, the parameter self-learning unit uses the measured coal flow rate measured by the belt scale during the current cycle. With model-predicted flow Parameter correction is performed to correct the error; The correction process is as follows: the predicted flow model is written as Define instantaneous error ; For the correction factor Proportional correction is applied: ,in To correct the step size factor; Uniaxial compressive strength of reference rock The reciprocal is corrected using gradient: ,in The learning rate; The corrected parameters are immediately fed back to the cutting resistance estimation and coal flow calculation stages, enabling continuous adaptive adjustment of the parameters.

[0016] The present invention has the following advantages over the prior art: By estimating the coal and rock cutting resistance coefficient online using the cutting motor current and traction speed of the tunneling machine and calculating the expected coal flow rate, the coal flow disturbance is pre-introduced into the belt conveyor speed control decision. A segmented speed control strategy based on predicted load trends directly outputs acceleration, deceleration, or hold commands without complex optimization solutions. This achieves dynamic balance between coal flow and conveying capacity, avoids coal stockpiling and light-load idling, and significantly reduces computational resource consumption. Combined with the transport capacity boundaries of multi-stage conveying equipment, collaborative constraints are formed to avoid downstream overload and ensure load balance across the entire transport line. The actual load torque is calculated using the drive motor current and dynamic equations and compared with the theoretical load torque residual to identify slippage and deviation faults. Low-cost and high-reliability real-time monitoring can be achieved without additional sensors. A laser coal flow profile scanner is used to correct the loose effect correction factor online using a sliding window averaging method, and the rock compressive strength and correction factor are updated incrementally step by step using coal weighing feedback. This allows the predicted coal flow rate to adaptively evolve with changes in coal and rock conditions, ensuring long-term stable prediction accuracy. This improves the overall operating efficiency, equipment life, and automation level of the belt conveyor system at the tunneling face. Attached Figure Description

[0017] Figure 1 This is a system block diagram of the present invention. Detailed Implementation

[0018] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0019] like Figure 1 As shown, the integrated intelligent control system for the belt conveyor at the tunneling face includes a sensing module, an edge computing module, and an execution module. The sensing module is used to collect real-time information on the cutting motor current and cutting arm traction speed of the tunneling machine, as well as the belt speed and coal distribution on the conveyor belt of the belt conveyor. The edge computing module deploys a collaborative optimization control algorithm based on a cut-and-transport coupled dynamic model. This algorithm is used to execute the following processes: The cutting motor current and cutting arm traction speed collected by the sensing module are used to estimate the coal and rock cutting resistance coefficient online. Based on the estimated coal and rock cutting resistance coefficient, cutting head geometric parameters and rotation speed, the expected coal flow rate in the future specified time domain is calculated; using the expected coal flow rate as a disturbance input, based on the preset dynamic energy consumption and wear model of the belt conveyor, a finite time domain optimization problem is constructed with the optimization objectives of conveying energy loss and conveyor belt wear rate, and the optimal conveyor belt speed reference trajectory is solved; The execution module generates speed control commands for the frequency converter of the belt conveyor based on the optimal conveyor belt speed reference trajectory, and outputs a traction speed suggestion value to the tunneling machine controller to achieve dynamic balance between coal cutting and conveying capacity.

[0020] Online estimation of the coal and rock cutting resistance coefficient is achieved through an extended Kalman filter based on the observation of the cutting motor load torque; The extended Kalman filter uses the coal and rock cutting resistance coefficient as a state variable and the cutting motor current and the cutting arm traction speed as measured values. It recursively calculates the estimated value of the coal and rock cutting resistance coefficient at the current moment through state prediction and measurement update. By utilizing the cutting motor current and cutting arm traction speed signals of the tunneling machine, the coal and rock cutting resistance coefficient is recursively estimated in real time as a state variable through an extended Kalman filter. This eliminates the need for additional coal and rock hardness sensors, allowing dynamic identification of the coal and rock resistance characteristics at the working face during the cutting process. This provides accurate and real-time updated input for subsequent calculation of the expected coal flow rate, overcoming the problem of large prediction deviations caused by traditional methods relying on fixed empirical values ​​or offline test data. It significantly improves the control system's ability to quickly adapt to changes in coal and rock conditions, enabling the belt conveyor speed adjustment decision to match sudden changes in cutting conditions in a timely manner, and avoiding coal pile-up or light-load idling caused by lag in coal flow rate prediction.

[0021] During the process of the tunneling machine cutting from the coal seam section into the semi-coal-rock section, the cutting resistance suddenly increases. The control system acquires the effective value of the cutting motor current I(k) and the cutting arm traction speed in each sampling cycle. .

[0022] Extended Kalman filter reduces the coal and rock cutting resistance coefficient Modeled as a random walk state, the state equation is: ,in The process noise has a covariance of .

[0023] The observation equation is established based on the torque balance relationship of the cutting motor: motor electromagnetic torque It equals the sum of the cutting resistance torque and the mechanical loss torque; the cutting resistance torque can be expressed as... ,in Here is the reference value for the uniaxial compressive strength of rock, and d(k) is the value determined by the traction speed. and cutting head speed The current cutting depth is determined, where R is the radius of the cutting head. It is a function related to the geometry and rotational speed of the cutting head, and the mechanical loss torque. Since the values ​​are known after offline calibration, the observation equation can be written as: ; in To observe the noise, its covariance is .

[0024] The filter recursion process is as follows: using the state estimate from the previous time step... Covariance Predicting the current prior estimate through the state equation Prior covariance ; Substitute the prior estimate into the observation equation to calculate the predicted current. And compare it with the actual sampled current I(k) to obtain the new information. ; Calculate the observation Jacobian matrix ; Kalman gain is Update posterior estimates Posterior covariance .

[0025] When the cutting target changes from coal to semi-coal-rock, the actual current I(k) increases significantly. If the Kalman gain is positive, the state estimate is corrected by weighting the new information, making... The rapid rise accurately reflects the sudden changes in the coal and rock resistance coefficient, ensuring the real-time and accurate prediction of subsequent coal flow rate.

[0026] The calculation process for the expected coal flow rate includes: Estimated coal and rock cutting resistance coefficient Cutting head radius R, cutting head rotation speed Based on the cutting depth d, the energy required for cutting a unit volume of rock is determined according to the cutting-to-energy ratio, and then the instantaneous coal flow rate is calculated by combining the real-time power of the cutting motor; the expected coal flow rate is then calculated. Obtained through the following formula: ; in, In order to cut the motor power, For reference, the uniaxial compressive strength of rock, The effective cutting area is related to the cutting head radius and cutting depth. A correction factor to account for the rock fracturing and loosening effect; The collaborative optimization control algorithm also maps the instantaneous volumetric flow rate to the expected coal drop flow rate sequence within the future time window based on the swing speed and cutting depth of the cutting arm; By using the extended Kalman filter to estimate the coal and rock cutting resistance coefficient in real time, combined with the cutting motor power, cutting head geometric parameters, and cutting depth, the instantaneous coal volume flow rate is directly calculated through the cutting specific energy consumption model. This eliminates the need to install dedicated coal quantity sensors on the cutting arm or chute. Furthermore, the instantaneous flow rate is mapped forward to a future coal flow rate sequence within a time window using the cutting arm swing speed. This provides accurate disturbance feedforward information for predictive speed regulation of the belt conveyor, enabling the conveying speed to respond in advance to changes in cutting conditions. This avoids coal pile-up or idling caused by delayed coal quantity detection, and improves the real-time performance and stability of the system's collaborative control.

[0027] During tunnel boring machine (TBM) excavation, the machine enters a semi-coal-rock section from the coal seam. At the current moment, the extended Kalman filter estimates the coal-rock cutting resistance coefficient as follows: Real-time power of the cutting motor Reference rock uniaxial compressive strength Cutting head radius Cutting head rotation speed The swaying traction speed of the cutting arm makes the current cutting depth d = 0.15m, and the loosening effect correction factor is corrected online to η = 1.3.

[0028] Effective cutting area Approximate calculation based on cutting head radius and cutting depth. .

[0029] Substitute these parameters into the formula for the expected coal flow rate: ; Calculate the denominator ,molecular The ratio is Multiply by the effective cross-sectional area and the looseness factor to get That is, the instantaneous coal drop volume flow rate is approximately The collaborative optimization control algorithm is based on the current swing speed of the cutting arm. Given the cutting depth d, calculate the time it will take for the cutting arm to sweep across the next section of the same resistance working face. Integrate the instantaneous flow rate on the time axis and map it forward to generate the expected coal drop flow rate sequence for the next 5 control cycles (0.5s per cycle). If the cutting conditions remain unchanged, the flow rate in each cycle will remain the same. Based on this sequence, the load change of the conveyor belt is predicted, and the speed of the belt conveyor is adjusted in advance to achieve balanced coal flow transportation.

[0030] The dynamic energy consumption and wear model of belt conveyors will be used to measure the total power of the belt conveyor. Expressed as a function of belt speed v and material load per unit length q: ; in, to The identification coefficients are related to the structural parameters of the conveyor and the characteristics of the idler rollers, where g is the acceleration due to gravity and H is the material lifting height. The conveyor belt wear rate W(v,q) is quantitatively calculated based on the contact stress between the conveyor belt and the idler rollers and the relative slip velocity. Its expression is as follows: ; in, and The wear coefficient is related to the material properties of the conveyor belt and idler rollers; Total power Together with the wear rate W(v,q), it forms the comprehensive quantitative basis for energy loss and mechanical wear in the conveying process; The total power of the belt conveyor and the wear rate of the conveyor belt are respectively constructed as analytical functions of belt speed and material load per unit length. By presetting identification coefficients related to the conveyor structural parameters and idler characteristics, as well as wear coefficients related to the conveyor belt material, online quantitative assessment of energy loss and mechanical wear during the conveying process is realized. No additional power monitoring and wear detection sensors are required. This provides an accurate and calculable comprehensive quantitative basis for constructing subsequent collaborative optimization control algorithms to address optimization problems aimed at energy saving, consumption reduction, and extending conveyor belt life. This enables speed adjustment decisions to directly balance conveying energy consumption and equipment wear, avoiding energy waste and excessive wear caused by speed adjustment based on experience.

[0031] During the operation of the belt conveyor at the tunneling face, the edge computing module calculates the material load per unit length on the conveyor belt based on the real-time coal distribution information obtained by the sensing module. Current belt speed of belt conveyors .

[0032] The conveyor's structural parameters are: lifting height H = 50m, gravitational acceleration... The coefficients obtained through offline identification Conveyor belt wear coefficient .

[0033] Substituting the above parameters into the total power formula, the calculation process is as follows: ; ; ; ; Power Boost ; Total power .

[0034] Simultaneously calculate the wear rate: .

[0035] The calculation results provide real-time quantitative values ​​of energy loss and mechanical wear under the current operating conditions, providing a direct basis for subsequent optimization algorithms to determine whether the belt speed needs to be adjusted. If it is predicted that the increase in coal flow will lead to an increase in q in the future, the total power and wear rate under different candidate belt speeds can be calculated in advance, and the speed command that makes the weighted objective function optimal can be selected.

[0036] The finite-time optimization problem in the collaborative optimization control algorithm is implemented using a piecewise speed regulation strategy based on predicted load trends. The specific process is as follows: Based on the projected coal flow rate sequence, the estimated unit length load at each time point in the future prediction time domain is obtained by recursively calculating the projected coal flow rate according to the conservation of material mass. and take its maximum value. ; when If the load exceeds the critical load corresponding to the maximum allowable filling rate of the conveyor belt at the current belt speed, and the overload trend continues to rise, the control belt conveyor should be adjusted from the current belt speed. To the predetermined high belt speed Smooth transition; when When the load is below the lower limit of the efficient operating load range at the current belt speed, and the light load trend continues to decrease, control the belt conveyor to reduce the load from the current belt speed. To the predetermined low-speed economic gear Smooth transition; Maintain the current belt speed in all other cases; During the belt speed transition, acceleration and deceleration are constrained by a preset ramp function, while ensuring that the real-time filling rate on the conveyor belt does not exceed the maximum allowable value; By recursively obtaining the maximum load per unit length in the future prediction time domain using the expected coal flow sequence, and comparing its trend with the allowable filling rate critical load and the high-efficiency operating load range at the current belt speed, segmented speed control commands for acceleration, deceleration, or maintenance are directly output. There is no need to deploy complex numerical optimization solvers in the edge computing module; the decision can be made simply by comparing the maximum load with the threshold value and judging the load trend direction. This greatly reduces the online computation and implementation complexity of the control algorithm. At the same time, by constraining acceleration and deceleration through the ramp function, the impact of sudden belt speed changes on the conveyor belt and drive system is avoided. Under the premise of ensuring dynamic balance between coal falling and conveying capacity and preventing coal piling and light-load idling, efficient and stable operation of speed control is achieved.

[0037] Current belt speed of belt conveyor at tunneling face The pre-set high-speed gear Low-speed economic gear The maximum allowable filling rate of the conveyor belt is 80%, corresponding to a critical load per unit length of [missing information]. The lower limit of the high-efficiency operating load range is .

[0038] The collaborative optimization control algorithm is based on the obtained predicted coal flow sequence. The predicted unit length load at each future time point is calculated recursively using the law of conservation of mass. The recursive relationship is as follows: ,in For the length of the conveyor belt carrying section, take... Assume the recursive load sequence per unit length for the next five control cycles is as follows: Take its maximum value This value exceeds the critical load. Furthermore, the load sequence shows a continuous upward trend. (Incrementing step by step) to meet the condition of a continuously rising overload trend, the segmented speed regulation strategy outputs from the current belt speed. Up to higher speed gear The acceleration command during the transition is constrained by a preset ramp function. The transition process continues During this period, the real-time filling rate never exceeds 80% of the maximum allowable filling rate, thereby accelerating the conveying speed in advance to cope with the upcoming coal flow peak and avoid coal pile-up accidents.

[0039] If, under another operating condition, the cutting rate of the tunneling machine decreases, the predicted unit length load sequence obtained by recursively calculating the expected coal flow rate sequence is: maximum value Although it did not exceed the critical load, it was below the lower limit of the high-efficiency operating load range. Furthermore, the load sequence shows a continuous downward trend. (Successively decreasing), satisfying the condition of a continuous decrease in light load trend, the segmented speed regulation strategy outputs from the current belt speed. Towards the low-speed gear of the economy The deceleration command during the transition is constrained by a preset ramp function. The transition process lasts approximately 6.7 seconds. While ensuring normal material transport, the belt speed is reduced to save energy and minimize equipment wear. If the predicted maximum load sequence value is... to If there is no continuous unidirectional trend between the two, then the current belt speed will remain unchanged to ensure the smooth operation of the conveyor system.

[0040] The edge computing module also receives material flow detection signals from the transfer machine or downstream belt conveyor, and transforms the transport capacity boundary of the downstream belt conveyor into additional constraints for the segmented speed regulation strategy, so as to realize speed coordination control between multi-level conveying equipment. The calculation process for transforming the transport capacity boundary of the downstream belt conveyor into constraint conditions is as follows: Based on the maximum permissible cross-sectional area of ​​the downstream belt conveyor Current real-time speed and material bulk density Calculate its maximum allowable transport flow rate. : ; When the estimated flow rate of the belt conveyor output to the transfer conveyor or downstream conveyor belt Exceed At that time, the segmented speed control strategy will output a speed reduction command to limit the speed. Not exceeding Output flow By predicting the load per unit length Obtained by multiplying by the conveyor belt speed; By receiving material flow detection signals from downstream belt conveyors, the maximum allowable material cross-sectional area, current belt speed, and material bulk density are converted into the maximum allowable conveying flow rate, which serves as an additional flow constraint for the segmented speed regulation strategy. When the predicted output flow rate exceeds the downstream receiving capacity, the speed of the upstream conveyor belt is actively limited. This achieves coordinated speed control among multi-level conveying equipment without increasing hardware costs, effectively preventing coal pile-up, material spillage, and equipment overload at transfer points caused by upstream material exceeding downstream carrying capacity. This ensures balanced load and continuous stable operation of the entire transportation system, and improves the reliability and automation level of multi-machine joint operations at the tunneling face.

[0041] In a multi-stage conveying system consisting of an upstream belt conveyor and a downstream conveyor belt of a transfer conveyor at the tunneling face, the current belt speed of the upstream conveyor is... The pre-set high-speed gear According to the segmented speed regulation strategy, the maximum load per unit length is predicted. Exceeding the critical load and continuing to rise, the proposed output direction The command to accelerate.

[0042] Simultaneously, the edge computing module receives real-time material flow detection signals from the downstream belt conveyor: its maximum permissible material cross-sectional area. Current speed Bulk density of materials Calculate the maximum allowable downstream flow rate: ; If the upstream belt speed directly increases to The predicted output flow rate is The value exceeds This will lead to coal accumulation at transfer points. The segmented speed control strategy imposes additional constraints: requiring... That is, the allowable upstream belt speed should meet the following requirements. Therefore, the speed adjustment strategy will accelerate the original plan to [speed limit]. The command was automatically corrected to accelerate only to And under the constraint of the ramp function, the transition is smooth, thereby maximizing the transmission efficiency while meeting the downstream receiving capacity, and avoiding operational accidents caused by speed mismatch of multiple machines.

[0043] The sensing module is also used to collect the current of the belt conveyor drive motor; the edge computing module also includes a conveyor belt slippage and misalignment monitoring unit, whose fault identification process is as follows: Based on the current signal and torque constant of the belt conveyor drive motor Calculate electromagnetic torque The calculation formula is: , where I is the effective value of the current calculated from the current signal; Based on the equivalent rotational inertia of the drive roller and the reducer The belt speed and drive drum radius of the belt conveyor Calculated roller angular velocity And the viscous friction coefficient B, using the driving drum dynamics equations to inversely calculate the actual load torque. The dynamic equation is: ; The estimated value of the actual load torque obtained by back-calculation is: ,in This refers to the angular acceleration of the drum. Based on the coal distribution information on the conveyor belt obtained by the sensing module, the mass of material per unit length is obtained. Combined with the mass per unit length of the conveyor belt Drive roller radius Conveyor inclination angle And the running resistance coefficient f, calculate the theoretical load torque caused by the weight of the material and the conveyor belt. The calculation formula is: ; Where g is the acceleration due to gravity; Calculate torque residual ; When torque residual If the absolute value exceeds the threshold calculated based on normal operating data and the duration exceeds the set confirmation time, it is determined that the conveyor belt has slipped or deviated. By utilizing the current signal from the drive motor of the belt conveyor and the coal distribution information acquired by the sensing module, the actual load torque is calculated back through the dynamic equation of the drive drum. This is then compared with the theoretical load torque calculated from the coal quantity and the self-weight of the conveyor belt. The amplitude and duration of the torque residual are used as criteria to identify slippage or misalignment faults. This eliminates the need for additional slippage detection switches or misalignment sensors, reducing system hardware costs and maintenance workload. Furthermore, the quantitative analysis based on the physical model can effectively distinguish between normal torque changes caused by load fluctuations and abnormal torque deviations caused by mechanical faults, improving the accuracy and timeliness of fault detection and providing a reliable guarantee for the safe operation of the conveyor belt.

[0044] The belt conveyor is in a stable operating state, and the torque constant of the drive motor is... The edge computing module collects the drive motor current signal and calculates the effective current value I = 60A. Then the electromagnetic torque... .

[0045] Equivalent moment of inertia of drive roller and reducer viscous friction coefficient Drive roller radius , by belt speed Calculate the roller angular velocity Angular acceleration under the current uniform motion state Calculate the actual load torque: ; The sensing module obtains the mass of material per unit length based on the coal quantity distribution information. Given the mass per unit length of the conveyor belt Conveyor inclination angle The drag coefficient f = 0.025, and the acceleration due to gravity. Calculate the theoretical load torque: ; in ,but Continue calculating .

[0046] Torque residual Its absolute value Exceeding the preset threshold If this state persists for more than the set confirmation time of 2 seconds, the monitoring unit determines that the conveyor belt has slipped. The actual load torque is significantly lower than the theoretical value, indicating that there is relative slippage between the drive roller and the conveyor belt, resulting in insufficient friction. The system then issues a slippage alarm signal. If a belt misalignment fault occurs, the additional frictional resistance caused by the belt deviation will typically result in a positive and continuously increasing residual, which can also be identified. This method utilizes existing signals for online calculation, eliminating the need for additional sensors and significantly improving the economy and practicality of fault monitoring.

[0047] The sensing module also includes a laser coal flow cross-section scanner; the edge computing module utilizes the measured coal flow volume retrieved by laser scanning. Compared with the expected coal flow rate Comparison of the loose effect correction factor Perform online calibration; The online calibration process is as follows: Collect a sequence of the ratios of the measured coal flow volume to the expected coal flow rate for the most recent L sampling periods. The arithmetic mean of the comparison value series is used to obtain the average ratio. : ; Use the current correction factor Updated to ; Where L is the sliding window length, the window length is selected according to the degree of coal flow fluctuation. When the fluctuation is severe, a smaller window is used for rapid response, and when the fluctuation is gentle, a larger window is used to smooth the noise. By utilizing the measured coal flow volume retrieved in real time using a laser coal flow cross-section scanner, a sliding window averaging correction is applied to the predicted coal flow rate. The loosening effect correction factor is updated by calculating the arithmetic mean of the ratios of measured and predicted values ​​over the most recent sampling periods. This effectively smooths out the random noise of a single laser scan and the measurement error caused by instantaneous fluctuations in the coal flow. At the same time, the window length is flexibly selected according to the degree of coal flow fluctuation, achieving a balance between response speed and correction stability. This allows the correction factor η to accurately track changes in the fracturing and loosening characteristics of coal and rock, thereby continuously improving the accuracy of the predicted coal flow rate. This provides a reliable predicted load input for the segmented speed regulation strategy and avoids speed regulation lag or malfunctions caused by η mismatch.

[0048] At the tunneling face, the current value of the loosening effect correction factor. The laser coal flow profile scanner of the sensing module in each sampling cycle Inverting the measured coal flow volume once The edge computing module simultaneously calculates the expected coal flow rate according to the formula. .

[0049] Within the most recent L = 5 sampling periods, the sequence of ratios between the measured coal flow volume and the expected coal flow rate is as follows: ; Among them, the expected coal flow rate It remains unchanged during periods of stable cutting conditions.

[0050] Take the arithmetic mean of this ratio sequence: ; According to the correction formula The updated correction factor is: ; Because the coal body is relatively loose after being broken, the actual coal flow volume is consistently larger than the expected value. A sliding window correction adjusts η from 1.3 to 1.454, correspondingly increasing the expected coal flow rate. Subsequent predictions will be closer to the actual coal flow. The segmented speed control strategy calculates the expected load based on the updated η, preparing for speed increases in advance and avoiding the risk of coal pile-up due to an underestimated predicted load caused by a small η. When the coal flow fluctuates drastically, the control system automatically shortens the window length to L = 3 to quickly track changes; when the coal flow is stable, the window is increased to L = 10 to further smooth noise and ensure the stability and accuracy of the correction.

[0051] The edge computing module further includes a parameter self-learning unit, which uses real-time feedback of coal weighing data during system operation to assess the uniaxial compressive strength of a reference rock. and loose effect correction factor Perform online incremental updates; At the end of each weighing cycle, the parameter self-learning unit uses the measured coal flow rate measured by the belt scale during the current cycle. With model-predicted flow Parameter correction is performed to correct the error; The correction process is as follows: the predicted flow model is written as Define instantaneous error ; For the correction factor Proportional correction is applied: ,in To correct the step size factor; Uniaxial compressive strength of reference rock The reciprocal is corrected using gradient: ,in The learning rate; The corrected parameters are immediately fed back to the cutting resistance estimation and coal flow calculation stages, realizing continuous adaptive adjustment of the parameters; By utilizing the measured coal flow rate fed back by the belt scale during the non-cutting interval of the tunneling machine or at the end of the weighing cycle, the uniaxial compressive strength and loosening effect correction factor of the reference rock are updated online incrementally. The parameters are gradually approximated to the true value by single-step error driving the parameters. There is no need to store a large amount of historical data or solve matrix equations. The computational load is extremely small and it can run continuously in the edge computing module. This allows the coal flow rate prediction model to automatically adapt to changes in the coal seam and rock strata interface. It overcomes the problem of the continuous decline in prediction accuracy of fixed parameter models when coal and rock conditions change frequently. It ensures the long-term accuracy of the predicted coal flow rate and the segmented speed regulation strategy, thereby improving the robustness and energy-saving effect of the system control.

[0052] The tunneling face has now transitioned from a semi-coal-rock section to a full-rock section, resulting in increased cutting resistance. The coal-rock cutting resistance coefficient has been updated using an extended Kalman filter. The cutting motor power has been increased to Cutting depth Effective cut area The current model parameters are Predicting traffic .

[0053] At the end of one weighing cycle, the measured average coal flow rate of the belt scale was: ,error relative error Take the correction step size coefficient. Update correction factor: ; Next, update the reference rock uniaxial compressive strength. Take the learning rate First calculate the current The gradient correction amount is: ; denominator ,molecular The fraction value is multiplied by gradient correction amount This value is so small as to be negligible (because...) Order of magnitude The reciprocal correction is not sensitive to a single error, which reflects the slow adjustment through self-learning.

[0054] After the actual update , Basically maintained .

[0055] After the parameters are updated, the predicted traffic becomes Closer to actual measurement The error is significantly reduced.

[0056] As subsequent weighing cycles are continuously adjusted, η and The values ​​gradually converge to match those of the entire rock section, ensuring that the coal flow prediction remains accurate.

[0057] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. An integrated intelligent control system for belt conveyors in tunneling faces, characterized in that, It includes a perception module, an edge computing module, and an execution module; The sensing module is used to collect real-time information on the cutting motor current and cutting arm traction speed of the tunneling machine, as well as the belt speed and coal distribution on the conveyor belt of the belt conveyor. The edge computing module deploys a collaborative optimization control algorithm based on a cut-and-transport coupled dynamic model. This algorithm is used to execute the following processes: The cutting motor current and cutting arm traction speed collected by the sensing module are used to estimate the coal and rock cutting resistance coefficient online. Based on the estimated coal and rock cutting resistance coefficient, cutting head geometric parameters and rotation speed, calculate the expected coal flow rate in the future specified time domain. Using the expected coal flow rate as the disturbance input, and based on a pre-set dynamic energy consumption and wear model of the belt conveyor, a finite-time domain optimization problem is constructed with the optimization objectives of conveying energy loss and conveyor belt wear rate, and the optimal conveyor belt speed reference trajectory is solved. The execution module generates speed control commands for the frequency converter of the belt conveyor based on the optimal conveyor belt speed reference trajectory, and outputs a traction speed suggestion value to the tunneling machine controller to achieve dynamic balance between coal falling and conveying capacity.

2. The integrated intelligent control system for belt conveyors in tunneling faces according to claim 1, characterized in that: Online estimation of the coal and rock cutting resistance coefficient is achieved through an extended Kalman filter based on the observation of the cutting motor load torque; The extended Kalman filter uses the coal and rock cutting resistance coefficient as a state variable and the cutting motor current and the cutting arm traction speed as measured values. Through state prediction and measurement update, it performs recursive calculations to obtain the estimated value of the coal and rock cutting resistance coefficient at the current moment.

3. The integrated intelligent control system for the belt conveyor in the tunneling face according to claim 2, characterized in that: The calculation process for the expected coal flow rate includes: Estimated coal and rock cutting resistance coefficient Cutting head radius R, cutting head rotation speed Based on the cutting depth d, the energy required for cutting a unit volume of rock is determined according to the cutting ratio energy consumption relationship, and then the instantaneous coal drop volume flow rate is calculated by combining the real-time power of the cutting motor. The collaborative optimization control algorithm also maps the instantaneous volumetric flow rate to the expected coal drop flow rate sequence within the future time window based on the swing speed and cutting depth of the cutting arm.

4. The integrated intelligent control system for the belt conveyor in the tunneling face according to claim 3, characterized in that: The dynamic energy consumption and wear model of the belt conveyor will determine the total power of the belt conveyor. It is expressed as a function of belt speed v and material load per unit length q; The wear rate of the conveyor belt is quantitatively calculated based on the contact stress between the conveyor belt and the idler roller and the relative slip velocity. Total power and wear rate together form the comprehensive quantitative basis for energy loss and mechanical wear in the conveying process.

5. The integrated intelligent control system for the belt conveyor in the tunneling face according to claim 4, characterized in that: The finite-time optimization problem in the collaborative optimization control algorithm is implemented using a piecewise speed regulation strategy based on predicted load trends. The specific process is as follows: Based on the projected coal flow rate sequence, the estimated unit length load at each time point in the future prediction time domain is obtained by recursively calculating the projected coal flow rate according to the conservation of material mass. and take its maximum value. ; when If the load exceeds the critical load corresponding to the maximum allowable filling rate of the conveyor belt at the current belt speed, and the overload trend continues to rise, the control belt conveyor should be adjusted from the current belt speed. To the predetermined high belt speed Smooth transition; when When the load is below the lower limit of the efficient operating load range at the current belt speed, and the light load trend continues to decrease, control the belt conveyor to reduce the load from the current belt speed. To the predetermined low-speed economic gear Smooth transition; Maintain the current belt speed in all other cases; During the belt speed transition, acceleration and deceleration are constrained by a preset ramp function, while ensuring that the real-time filling rate on the conveyor belt does not exceed the maximum allowable value.

6. The integrated intelligent control system for belt conveyors in tunneling faces according to claim 5, characterized in that: The edge computing module also receives material flow detection signals from the transfer machine or downstream belt conveyor, and transforms the transport capacity boundary of the downstream belt conveyor into additional constraints for the segmented speed regulation strategy, so as to realize the speed coordination control between multi-level conveying equipment. The calculation process for transforming the transport capacity boundary of the downstream belt conveyor into constraint conditions is as follows: Based on the maximum permissible cross-sectional area of ​​the downstream belt conveyor Current real-time speed and material bulk density Calculate its maximum allowable transport flow rate. ; When the estimated flow rate of the belt conveyor output to the transfer conveyor or downstream conveyor belt Exceed At that time, the segmented speed control strategy will output a speed reduction command to limit the speed. Not exceeding Output flow By predicting the load per unit length It is obtained by multiplying the speed of the conveyor belt.

7. The integrated intelligent control system for belt conveyors in tunneling faces according to claim 6, characterized in that: The sensing module is also used to collect the current of the belt conveyor drive motor; the edge computing module also includes a conveyor belt slippage and misalignment monitoring unit, whose fault identification process is as follows: Based on the current signal and torque constant of the belt conveyor drive motor Calculate electromagnetic torque ; Based on the equivalent rotational inertia of the drive roller and the reducer The belt speed and drive drum radius of the belt conveyor Calculated roller angular velocity And the viscous friction coefficient B, using the dynamic equations of the driving drum to inversely calculate the actual load torque. ; Based on the coal distribution information on the conveyor belt obtained by the sensing module, the mass of material per unit length is obtained. Combined with the mass per unit length of the conveyor belt Drive roller radius Conveyor inclination angle And the running resistance coefficient f, calculate the theoretical load torque caused by the weight of the material and the conveyor belt. , ; Calculate torque residual When torque residual If the absolute value exceeds the threshold calculated based on normal operating data and the duration exceeds the set confirmation time, it is determined that the conveyor belt has slipped or deviated.

8. The integrated intelligent control system for the belt conveyor in the tunneling face according to claim 7, characterized in that: The sensing module also includes a laser coal flow cross-section scanner; the edge computing module utilizes the measured coal flow volume retrieved by laser scanning. Compared with the expected coal flow rate Comparison of the loose effect correction factor Perform online calibration; The online calibration process is as follows: Collect a sequence of the ratios of the measured coal flow volume to the expected coal flow rate for the most recent L sampling periods. The arithmetic mean of the comparison value series is used to obtain the average ratio. ; Use the current correction factor Updated to ; Where L is the sliding window length, the window length is selected according to the degree of coal flow fluctuation. When the fluctuation is severe, a smaller window is used for rapid response, and when the fluctuation is gentle, a larger window is used to smooth the noise.

9. The integrated intelligent control system for belt conveyors in tunneling faces according to claim 8, characterized in that: The edge computing module further includes a parameter self-learning unit, which uses real-time feedback of coal weighing data during system operation to assess the uniaxial compressive strength of a reference rock. and loose effect correction factor Perform online incremental updates; At the end of each weighing cycle, the parameter self-learning unit uses the measured coal flow rate measured by the belt scale during the current cycle. With model-predicted flow Parameter correction is performed to correct the error; The correction process is as follows: the predicted flow model is written as Define instantaneous error ; For correction factor Use proportional correction; Uniaxial compressive strength of reference rock The reciprocal of the equation is corrected using gradient descent to obtain the corrected parameters; The corrected parameters are immediately fed back to the cutting resistance estimation and coal flow calculation stages, enabling continuous adaptive adjustment of the parameters.

Images