Long-distance complex working condition belt conveyor multi-mode cooperative variable frequency control method and system

Abstract

The application discloses a long-distance complex working condition belt conveyor multi-mode cooperative variable frequency control method and system, and belongs to the technical field of bulk material conveying automation and intelligent control. The method comprises the following steps: S1, a digital twin prior model construction step; S2, a multi-mode starting control step; S3, a multi-machine asymmetric power balance control step; S4, an energy-saving cruise step based on flow prediction; and S5, a multi-target cooperative braking step. The application is aimed at the complex field working conditions of long-distance belt conveyor variable gradient, variable load and multi-point driving. Relying on digital twin prediction of belt deformation, the application is coupled with deviation decoupling, flow prediction speed regulation and electrical-mechanical composite braking, effectively suppresses start-stop tension fluctuation, eliminates parasitic power and oscillation among multi-motors, reduces equipment mechanical loss, simultaneously realizes advanced speed regulation and energy saving according to incoming materials, extends control protection from traditional electrical protection to active intervention of belt mechanical state, and improves the running stability of the whole machine and the safety of field operation.

Description

Technical Field

[0001] This invention relates to the field of automation and intelligent control technology for bulk material conveying, and more specifically, to a multi-modal collaborative frequency conversion control method and system for belt conveyors under long-distance and complex working conditions. Background Technology

[0002] In bulk material handling scenarios such as mines, ports, and tunnels, long-distance belt conveyors are the core equipment for material transportation, with some single units capable of conveying lengths exceeding 5km. To meet the demands of heavy-load conveying, these belt conveyors generally employ multiple motors at the head, tail, or middle points for joint drive. During on-site operation, they face practical challenges such as undulating slopes along the route, unstable material flow, and significant elastic deformation of the belt itself. Dynamic tension and power matching among multiple motors during equipment operation have always been key challenges in on-site commissioning and maintenance.

[0003] Patent CN122610874A discloses a scheduling method, scheduling system, electronic device, and storage medium, belonging to the field of port scheduling technology. The scheduling method includes: acquiring the stopping times of all trains arriving at the port within a target time period, information on all tippers, belt conveyors, stackers, and pallet positions at the port; based on the train stopping times, information on all tippers, belt conveyors, stackers, and pallet positions, formulating multiple work plans for all trains to obtain an initial work plan set; determining the optimal work plan from the initial work plan set, and scheduling according to the optimal work plan. This application can efficiently determine the optimal work plan and schedule according to it, improving scheduling efficiency.

[0004] While this method offers numerous advantages, most existing belt conveyor control schemes rely on PID regulation and current-synchronized master-slave control, which only enables basic equipment start-up and shutdown and speed tracking. When encountering sudden load changes, the control parameters cannot adaptively adjust to the on-site conditions, resulting in significant speed lag and a tendency for belt resonance and motor stalling. Furthermore, traditional torque balancing methods struggle to eliminate the mutual push-pull forces generated by the flexible coupling of the belt, which can exacerbate fatigue damage to transmission components such as the reducer and coupling over long-term operation. Summary of the Invention

[0005] (a) Technical problems to be solved

[0006] To address the shortcomings of existing technologies, this invention provides a multi-modal collaborative frequency conversion control method and system for long-distance, complex-condition belt conveyors, thus solving the aforementioned problems.

[0007] (II) Technical Solution

[0008] To achieve the above-mentioned objectives, the present invention provides the following technical solution: a multi-modal collaborative frequency conversion control method and system for long-distance complex working conditions belt conveyors, applied to a belt conveyor control system configured with a main controller, multiple frequency converters, multiple drive motors, a hydraulic braking system, and a multi-source sensor network; the method includes the following steps:

[0009] S1 Digital twin prior model construction steps, S2 Multimodal start-up control steps, S3 Multi-machine asymmetric power balance control steps, S4 Energy-saving cruise based on flow prediction steps, S5 Multi-target cooperative braking steps;

[0010] The specific steps for constructing the S1 digital twin prior model are as follows: a digital twin of the belt conveyor built based on a distributed parameter model is pre-set inside the main controller; electrical parameters of the belt conveyor are entered into the digital twin of the belt conveyor; the Kelvin-Voigt belt viscoelastic mechanical model is embedded at the same time; field detection data uploaded in real time by a multi-source sensor network is received; and a corresponding mapping relationship between belt deformation and belt running speed under different belt tension conditions is established.

[0011] The S2 multimodal start-up control steps include a warm-up and self-test process, a preload establishment process, and an S-shaped adaptive acceleration process. After receiving the belt conveyor start-up command, the system enters the warm-up and self-test process, reading locally stored belt conveyor downtime data and ambient temperature data collected on-site. A fuzzy inference algorithm is used to output a motor heating command or a motor cooling command. After completing the warm-up and self-test process, the system enters the preload establishment process. The main controller sends excitation commands to all frequency converters, switching them to torque control mode. The system continuously collects the measured total output torque value and compares it with the minimum anti-backslip critical torque calculated using the belt's self-weight and the belt conveyor's laying slope. If the measured total output torque value is greater than the minimum anti-backslip critical torque, the system enters the S-shaped adaptive acceleration process. The S-shaped adaptive acceleration process reads the belt elastic deformation limit parameters output by the digital twin in step S1, generates a variable-parameter S-shaped speed curve, reduces the acceleration value for sections with large belt sag, and increases the acceleration value for sections with low belt tension, thus constraining the peak tension throughout the belt start-up process.

[0012] Preferably, the S1 digital twin prior model construction step further includes a model initialization and calibration sub-step; the model initialization sub-step imports the physical parameters of the belt entity into the main controller, including the belt's mass per unit length, belt elastic modulus, belt cross-sectional area, and elevation of each measuring point along the belt conveyor; before each execution of the S2 multimodal start control step, the measured temperature data of the entire line collected by the distributed fiber optic temperature measurement system is read, and the measured temperature data of the entire line is used to correct the belt friction coefficient stored inside the digital twin, reducing the deviation between the digital twin simulation output data and the actual operating data of the belt conveyor on site; the digital twin stores the basic parameters of the motor heat conduction at each drive point for parameter substitution calculation under abnormal sensor conditions.

[0013] Preferably, in the pre-tensioning force establishment process of the S2 multi-modal start-up control step, the frequency converter outputs a fixed low-frequency voltage to establish the motor excitation magnetic field, continuously collects the system output torque, compares the system output torque with the static resistance torque of the belt conveyor, and when the system output torque is greater than the static resistance torque, it is determined that the torque output can overcome the static friction resistance of the whole machine, and the main controller outputs a signal to release the brake lock of the hydraulic braking system; the S-shaped adaptive acceleration process of the S2 step pre-writes the maximum acceleration limit of the whole machine, calls the belt sag function output by the digital twin, configures the sag influence coefficient, the larger the output value of the belt sag function, the smaller the value of the sag influence coefficient, and the corresponding section's actual running acceleration decreases synchronously; the smaller the output value of the belt sag function, the larger the value of the sag influence coefficient, and the corresponding section's actual running acceleration increases synchronously.

[0014] Preferably, after completing the S2 multi-modal start-up control step and the belt conveyor enters a stable operating state, the S3 multi-machine asymmetric power balance control step is executed. The S3 step abandons the average current equalization adjustment method and adopts a deviation coupling control strategy to complete the decoupling of multiple drive motors. The actual speed and actual torque of each drive motor are collected according to the set sampling period. The system average torque is calculated based on the measured torque values ​​of all drive motors. The torque deviation of each axis between the measured torque value of each drive motor and the system average torque is calculated. A cross-coupling compensator is configured, and the position weights of each drive motor corresponding to the topology position of the belt conveyor are entered. The position weight of the head drive motor is greater than that of the tail drive motor. The torque setpoint of each motor is calculated by combining the torque deviation of each axis and the position weight. The torque setpoint is sent to the corresponding strain frequency converter to counteract the mutual pushing and pulling internal forces of the motors generated by the elastic expansion and contraction spring effect of the belt.

[0015] Preferably, in the S3 multi-machine asymmetric power balance control step, the cross-coupling compensator has a preset cross-coupling coefficient. In each sampling period, the frequency setpoint of each inverter is corrected by using the torque deviation, position weight, and cross-coupling coefficient of each drive motor. When any one of the drive motors experiences load disturbance fluctuation, the cross-coupling compensator generates a compensation component and distributes it to all the other drive motors. The load disturbance is shared by multiple motors, suppressing the whole machine oscillation caused by sudden load changes of a single motor.

[0016] Preferably, while executing the S3 multi-machine asymmetric power balance control step, the S4 energy-saving cruise step based on flow prediction is executed in parallel. Radar level gauges and AI vision acquisition devices are deployed at each material transfer point of the belt conveyor. The radar level gauges and AI vision acquisition devices synchronously collect incoming material information and predict the material flow rate entering the belt within the next T seconds based on the incoming material information. The current actual belt speed of the belt conveyor is read, and the optimal economic belt speed of the belt is calculated in combination with the material flow prediction result. The generalized predictive control algorithm is called, and the optimal economic belt speed is used as the control target to modify the output frequency of each frequency converter in advance, so that the actual belt speed of the belt smoothly follows the optimal economic belt speed, reducing the frequent acceleration and deceleration actions caused by sudden changes in material flow.

[0017] Preferably, in step S4, the energy-saving cruise step based on flow prediction establishes the belt conveyor dynamic equation. The state vector of the dynamic equation includes the belt running speed and the belt internal tension, and the control vector of the dynamic equation includes the inverter output torque. In each control cycle, the weighing data collected by the on-site weighing sensor is read, and the material distribution of the entire belt in the next 10 seconds is predicted based on the weighing data. The belt slippage prevention and belt overflow prevention are set as constraint boundaries, and the quadratic programming problem is solved to obtain the optimal motor torque sequence. The control quantity corresponding to the optimal motor torque sequence is sent to each inverter to complete the speed regulation.

[0018] Preferably, upon receiving a belt conveyor stop command, the S5 multi-target coordinated braking step is executed. The S5 multi-target coordinated braking step adopts a composite braking mode combining electrical braking and mechanical braking. Electrical braking is prioritized, with each frequency converter performing DC braking and energy-consuming braking. The frequency converter switches to power generation mode to feed the potential energy of the belt conveyor back to the grid, while simultaneously outputting braking torque. The anti-slip condition is checked in conjunction with the belt slope parameters, and the mechanical kinetic energy of the entire machine is consumed under the premise of meeting the anti-slip condition. The drive motor speed and frequency converter output current are continuously collected. When the drive motor speed drops to the preset critical speed and the frequency converter output current is lower than the preset safe current threshold, the main controller outputs a command to trigger the hydraulic braking system to hydraulically hold the brake to complete the mechanical braking intervention.

[0019] Preferably, during the operation of the whole machine, if the temperature sensor of a certain drive motor collects an abnormal signal, the main controller blocks the collection signal of the abnormal sensor, retrieves the heat conduction simulation data of the motor at the same location obtained from the S1 digital twin prior model construction step to replace the abnormal collection signal, and at the same time issues a frequency reduction operation command to each frequency converter, so that the belt conveyor can maintain continuous conveying operation and achieve non-stop maintenance.

[0020] This multi-modal collaborative frequency conversion control system for long-distance, complex-condition belt conveyors includes a main controller unit, a frequency conversion drive unit, a drive motor unit, a hydraulic braking unit, and a multi-source sensor acquisition unit. The main controller unit employs a high-performance redundant PLC and internally loads a digital twin program module, a fuzzy inference calculation module, a deviation coupling control calculation module, a generalized predictive control calculation module, a quadratic programming solution module, and a braking logic scheduling module. The frequency conversion drive unit comprises multiple four-quadrant frequency converters, each electrically connected to a permanent magnet synchronous motor within the drive motor unit. The hydraulic braking unit is equipped with multiple sets of hydraulic disc brakes, with the controlled end of the hydraulic disc brakes connected to the output signal lines of the main controller unit. The multi-source sensor acquisition unit includes a weighing sensor, a radar level gauge, an AI vision acquisition component, a distributed fiber optic temperature measurement device, a vibration monitor, a speed detection element, a torque detection element, and an ambient temperature detection element. The weighing sensor is installed on the frame of the conveyor belt carrying section. The radar level gauge and the AI ​​vision acquisition component are installed at each material transfer point. The distributed fiber optic temperature measurement device is laid along the entire conveyor belt. The vibration monitor is fixedly installed on the surface of the drive motor and the reducer housing. The speed detection element and the torque detection element are respectively assembled on the output shaft of each drive motor. The ambient temperature detection element is arranged at the conveyor belt operation site. All signal output terminals of the multi-source sensor acquisition unit are connected to the signal input port of the main controller unit.

[0021] (III) Beneficial Effects

[0022] Compared with the prior art, the present invention provides a multi-modal cooperative frequency conversion control method and system for long-distance complex working conditions belt conveyors, which has the following beneficial effects:

[0023] This invention addresses the complex on-site conditions of long-distance belt conveyors with varying slopes, loads, and multi-point drives. It relies on digital twins to predict belt deformation, combined with deviation coupling decoupling, flow prediction speed regulation, and electromechanical composite braking. This effectively suppresses tension fluctuations during start-stop operations, eliminates parasitic power and oscillations between multiple motors, and reduces mechanical losses. At the same time, it achieves advanced speed regulation and energy saving based on incoming materials, extending control and protection from traditional electrical protection to proactive intervention in the mechanical state of the belt, thereby improving the overall stability of the machine and the safety of on-site operations. Attached Figure Description

[0024] Figure 1 This is a flowchart of the present invention;

[0025] Figure 2 This is a framework diagram of the present invention. Detailed Implementation

[0026] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0027] Please see Figures 1-2 The present invention provides the following technical solution:

[0028] A multi-modal coordinated frequency conversion control method for long-distance, complex-condition belt conveyors, comprising the following steps:

[0029] The steps are executed sequentially: S1 Digital Twin Prior Model Construction, S2 Multimodal Start-up Control, S3 Multi-machine Asymmetric Power Balance Control, S4 Energy-saving Cruise Based on Flow Prediction, and S5 Multi-target Cooperative Braking.

[0030] A digital twin of the belt conveyor, built based on a distributed parameter model, is pre-installed inside the redundant PLC main controller. The digital twin records all electrical parameters of the belt conveyor and embeds the Kelvin-Voigt belt viscoelastic mechanical model.

[0031] The model initialization sub-step is executed to import the physical parameters of the belt entity into the main controller: belt mass per unit length, belt elastic modulus, belt cross-sectional area, and elevation data of each measuring point along the entire belt conveyor line, thus completing the basic model construction; the digital twin pre-stores the basic parameters of the heat conduction of the motors at each drive point at the head and tail, which are used for parameter replacement calculations in case of subsequent sensor anomalies.

[0032] Before each belt conveyor start-up operation, the actual temperature data of the entire belt line collected by the distributed fiber optic temperature measurement equipment is read. The actual temperature data is used to iteratively correct the belt friction coefficient stored in the digital twin, eliminating model deviations caused by changes in ambient temperature, so that the deformation and tension values ​​output by the digital twin simulation match the actual operating state of the belt on site. The digital twin receives real-time uploads of speed, torque, temperature, and weighing detection data from a multi-source sensor network, and establishes a mapping relationship between belt deformation and running speed under different tension conditions, providing simulation prediction basis for subsequent start-up acceleration, power decoupling, speed regulation, and braking.

[0033] The S2 process is divided into a warm-up and self-inspection process, a preload establishment process, and an S-shaped adaptive acceleration process.

[0034] After receiving the start command for the entire conveyor belt from the on-site control panel, the system first enters the warm-up and self-test process. The main controller retrieves the downtime record of the last shutdown from its local storage and simultaneously reads the measured ambient temperature value collected by the on-site ambient temperature detection element. The downtime and ambient temperature are input into the fuzzy inference algorithm module. When the ambient temperature is low and the downtime is long, a motor heating command is output to preheat the four permanent magnet synchronous motors. When the on-site ambient temperature is too high and the downtime interval is short, a motor cooling command is output to avoid the problem of excessive viscous resistance when the motor starts at low temperatures or overload when starting directly at high temperatures.

[0035] After completing the warm-up and self-test, the preload establishment process begins. The main controller sends excitation commands to all four frequency converters, which switch to torque control mode and output a 5Hz low-frequency voltage to establish the internal excitation magnetic field of the motor. The system continuously collects the measured value of the total output torque and compares it with the minimum anti-backslip critical torque calculated based on the belt weight and the belt conveyor slope, as well as the static resistance torque of the entire machine. When the measured value of the total output torque is greater than the minimum anti-backslip critical torque and also greater than the static resistance torque of the entire machine, it is determined that the output torque can overcome the static friction resistance of the entire machine. The main controller outputs a control signal to release the hydraulic disc brake from its locked state, and the process enters the S-shaped adaptive acceleration process.

[0036] In the S-shaped adaptive acceleration process, the main controller reads the belt elastic deformation limit parameters output by the S1 digital twin and simultaneously calls the belt sag function output by the digital twin to preset the maximum acceleration limit of the entire machine, introducing a sag influence coefficient. In sections with large sag function output values, such as the midpoint of the belt span, the sag influence coefficient is reduced, resulting in a decrease in the actual acceleration at the corresponding position and suppressing belt vibration. In the head and tail sections with higher tension, the sag function output value is smaller, the sag influence coefficient is increased, and the acceleration in the corresponding section is increased, balancing startup efficiency. Based on the above rules, a variable parameter S-shaped speed curve is generated. The frequency converter is controlled to increase the output frequency according to this curve, constraining the peak dynamic tension of the belt throughout the entire startup acceleration process and reducing belt stress impact during the startup phase.

[0037] After the belt conveyor completes startup and enters stable operation, the S3 machine asymmetric power balance control step is executed. In this embodiment, the traditional average current equalization adjustment method is abandoned, and the deviation coupling control strategy is adopted to achieve decoupled control of the four drive motors.

[0038] The system is set to a fixed sampling period of 20ms. At each sampling period node, the speed detection element and torque detection element synchronously collect the actual speed and actual torque of the four motors M1, M2, M3, and M4 and upload them to the main controller. The main controller calculates the system average torque based on the measured torque values ​​of the four motors and calculates the torque deviation of each axis between the measured torque value of each motor and the system average torque.

[0039] The cross-coupling compensator has a preset cross-coupling coefficient and records the topological position weights of each drive motor. The position weights of the head drive motors M1 and M2 are set to 0.6, and the position weights of the tail drive motors M3 and M4 are set to 0.4. The torque setpoint and frequency setpoint of each inverter are corrected by combining the torque deviation of each shaft, position weights, and cross-coupling coefficients, and then sent to the corresponding four-quadrant inverters.

[0040] When fluctuations in incoming materials or changes in slope cause load disturbance fluctuations in any of the motors, the cross-coupling compensator generates a corresponding compensation component and distributes it to the other three drive motors. The load disturbance is shared by all four motors, eliminating the high-frequency push-pull internal force between motors caused by the elastic extension and contraction of the belt, suppressing the mechanical vibration of the whole machine, and reducing the fatigue wear of the reducer and coupling.

[0041] While S3 is executing the operation control, the system runs the energy-saving cruise step S4 based on flow prediction in parallel.

[0042] At each transfer point of the belt conveyor, radar level gauges and AI vision acquisition components simultaneously collect incoming material information such as material layer height and cross-sectional shape, predicting the material flow rate entering the belt within the next T seconds (T is taken as 10s). The main controller reads the real-time actual belt speed and instantaneous material weight collected by the weighing sensor, and combines it with the predicted material flow rate to calculate the optimal economic belt speed target value for the belt conveyor.

[0043] The system constructs the belt conveyor's dynamic equations. The state vector of the dynamic equations includes the belt's running speed and internal tension, while the control vector includes the inverter's output torque. Each control cycle reads real-time weighing data collected by the load cells on the load-bearing section frame and predicts the material distribution along the entire belt over the next 10 seconds. Belt slippage and material overflow are set as constraint boundary conditions. A generalized predictive control algorithm is invoked to solve a quadratic programming (QP) problem, yielding the optimal motor torque sequence that satisfies the constraints. The control quantity corresponding to the optimal motor torque sequence is then sent to each inverter to pre-adjust the inverter's output frequency, ensuring the actual belt speed smoothly follows the optimal economic speed and avoiding frequent acceleration and deceleration caused by sudden changes in material flow. Under low material load conditions, the belt automatically reduces its operating speed, achieving energy savings while maintaining conveying capacity.

[0044] When a stop command is received from the control panel or host computer, the entire machine enters the S5 multi-target coordinated braking step, adopting a composite braking mode that prioritizes electrical braking and lags behind mechanical braking.

[0045] First, the electric braking process is executed. The four frequency converters switch to the power generation mode and simultaneously perform DC braking and energy consumption braking. The gravitational potential energy of the belt conveyor is fed back to the power grid through the motor and frequency converter, while continuously outputting braking torque. The control logic continuously verifies the anti-backslip constraint conditions under large tilt angle conditions, and consumes the mechanical kinetic energy of the whole machine while satisfying the anti-backslip condition.

[0046] The system continuously collects the real-time speed of the drive motor and the output current of the frequency converter. In this embodiment, the critical speed threshold is set at 50 rpm, and the safe current threshold is set at 5% of the motor's rated current. When the drive motor speed is detected to drop below 50 rpm, and the frequency converter output current is lower than the safe current threshold, it means that the motor is no longer outputting effective drive / braking torque. The main controller outputs a trigger command to control the hydraulic disc brake to engage the brake, completing the mechanical braking intervention.

[0047] During normal operation of the whole machine, a fault self-healing logic is set: if any of the four motors' associated temperature sensors experience abnormal acquisition signals such as disconnection or value jumps, the main controller directly blocks the acquisition signal of the abnormal sensor and retrieves the heat conduction simulation data of the motor at the same location in the S1 digital twin prior model as temperature substitute data to participate in the monitoring calculation; at the same time, it issues a frequency reduction operation command to all frequency converters, and the belt conveyor reduces speed to maintain uninterrupted conveying operation, so that sensor maintenance can be carried out on site without production interruption.

[0048] This embodiment corresponds to a multi-modal collaborative frequency conversion control system for long-distance, complex-condition belt conveyors, including a main controller unit, a frequency conversion drive unit, a drive motor unit, a hydraulic braking unit, and a multi-source sensor acquisition unit. The main controller unit is a high-performance redundant PLC, internally loaded with a digital twin program module, a fuzzy inference calculation module, a deviation coupling control calculation module, a generalized predictive control calculation module, a quadratic programming solution module, and a braking logic scheduling module. The frequency conversion drive unit contains four four-quadrant frequency converters, each electrically connected to one of the four permanent magnet synchronous drive motors within the drive motor unit. The hydraulic braking unit is equipped with multiple sets of hydraulic disc brakes, with the controlled end of the hydraulic disc brakes electrically connected to the signal output end of the main controller unit. The multi-source sensing and acquisition unit includes a load cell, a radar level gauge, an AI vision acquisition component, a distributed fiber optic temperature measurement device, a vibration monitor, a speed detection element, a torque detection element, and an ambient temperature detection element. The load cell is installed on the frame of the conveyor belt carrying section. The radar level gauge and the AI ​​vision acquisition component are installed at each material transfer point. The distributed fiber optic temperature measurement device is laid along the entire 5km conveyor belt. The vibration monitor is fixed on the surface of the drive motor and reducer housing. The speed detection element and torque detection element are assembled on the output shaft of each drive motor. The ambient temperature detection element is arranged on the conveyor belt site. The signal output terminals of all sensing devices are connected to the signal input port of the main controller unit.

[0049] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A multi-modal cooperative frequency conversion control method for long-distance, complex-condition belt conveyors, characterized in that: A belt conveyor control system, applicable to a system comprising a main controller, multiple frequency converters, multiple drive motors, a hydraulic braking system, and a multi-source sensor network, is described; the method includes the following steps: S1 Digital twin prior model construction steps, S2 Multimodal start-up control steps, S3 Multi-machine asymmetric power balance control steps, S4 Energy-saving cruise based on flow prediction steps, S5 Multi-target cooperative braking steps; The specific steps for constructing the S1 digital twin prior model are as follows: a digital twin of the belt conveyor built based on a distributed parameter model is pre-set inside the main controller; electrical parameters of the belt conveyor are entered into the digital twin of the belt conveyor; the Kelvin-Voigt belt viscoelastic mechanical model is embedded at the same time; field detection data uploaded in real time by a multi-source sensor network is received; and a corresponding mapping relationship between belt deformation and belt running speed under different belt tension conditions is established. The S2 multimodal start-up control steps include a warm-up and self-test process, a preload establishment process, and an S-shaped adaptive acceleration process. After receiving the belt conveyor start-up command, the system enters the warm-up and self-test process, reading locally stored belt conveyor downtime data and ambient temperature data collected on-site. A fuzzy inference algorithm is used to output a motor heating command or a motor cooling command. After completing the warm-up and self-test process, the system enters the preload establishment process. The main controller sends excitation commands to all frequency converters, switching them to torque control mode. The system continuously collects the measured total output torque value and compares it with the minimum anti-backslip critical torque calculated using the belt's self-weight and the belt conveyor's laying slope. If the measured total output torque value is greater than the minimum anti-backslip critical torque, the system enters the S-shaped adaptive acceleration process. The S-shaped adaptive acceleration process reads the belt elastic deformation limit parameters output by the digital twin in step S1, generates a variable-parameter S-shaped speed curve, reduces the acceleration value for sections with large belt sag, and increases the acceleration value for sections with low belt tension, thus constraining the peak tension throughout the belt start-up process.

2. The multi-modal cooperative frequency conversion control method for long-distance complex working conditions belt conveyors according to claim 1, characterized in that, The S1 digital twin prior model construction step also includes a model initialization and calibration sub-step. The model initialization sub-step imports the physical parameters of the belt entity into the main controller. The imported physical parameters include the belt's mass per unit length, belt elastic modulus, belt cross-sectional area, and elevation of each measuring point along the belt conveyor. Before each execution of the S2 multimodal start control step, the measured temperature data of the entire line collected by the distributed fiber optic temperature measurement system is read. The measured temperature data of the entire line is used to correct the belt friction coefficient stored inside the digital twin, reducing the deviation between the digital twin simulation output data and the actual operating data of the belt conveyor on site. The digital twin stores the basic parameters of the motor heat conduction at each drive point for parameter substitution calculations under abnormal sensor conditions.

3. The multi-modal cooperative frequency conversion control method for long-distance complex working conditions belt conveyors according to claim 1, characterized in that, In the pre-tensioning process of the S2 multi-modal start-up control step, the inverter outputs a fixed low-frequency voltage to establish the motor excitation magnetic field, continuously collects the system output torque, and compares the system output torque with the static resistance torque of the belt conveyor. When the system output torque is greater than the static resistance torque, it is determined that the torque output can overcome the static friction resistance of the whole machine, and the main controller outputs a signal to release the brake lock of the hydraulic braking system. In the S2 step, the S-shaped adaptive acceleration process pre-writes the maximum acceleration limit of the whole machine, calls the belt sag function output by the digital twin, and configures the sag influence coefficient. The larger the output value of the belt sag function, the smaller the value of the sag influence coefficient, and the corresponding section's actual running acceleration decreases synchronously; the smaller the output value of the belt sag function, the larger the value of the sag influence coefficient, and the corresponding section's actual running acceleration increases synchronously.

4. The multi-modal cooperative frequency conversion control method for long-distance complex working conditions belt conveyors according to claim 1, characterized in that, After completing the S2 multi-modal start-up control step and the belt conveyor enters a stable operating state, the S3 multi-motor asymmetric power balance control step is executed. The S3 step abandons the average current equalization adjustment method and adopts a deviation coupling control strategy to complete the decoupling of multiple drive motors. The actual speed and actual torque of each drive motor are collected according to the set sampling period. The system average torque is calculated based on the measured torque values ​​of all drive motors. The torque deviation of each axis between the measured torque value of each drive motor and the system average torque is calculated. A cross-coupling compensator is configured, and the position weights of each drive motor corresponding to its position on the belt conveyor topology are entered. The position weight of the head drive motor is greater than that of the tail drive motor. The torque setpoint of each motor is calculated by combining the torque deviation of each axis and the position weight. The torque setpoint is sent to the corresponding strain gauge frequency converter to counteract the mutual pushing and pulling internal forces of the motors generated by the elastic expansion and contraction spring effect of the belt.

5. The multi-modal cooperative frequency conversion control method for long-distance complex working conditions belt conveyors according to claim 4, characterized in that, In the S3 multi-machine asymmetric power balance control step, the cross-coupling compensator has a preset cross-coupling coefficient. In each sampling period, the frequency setpoint of each inverter is corrected by using the torque deviation, position weight, and cross-coupling coefficient of each drive motor. When any one of the drive motors experiences load disturbance fluctuation, the cross-coupling compensator generates a compensation component and distributes it to all the other drive motors. The load disturbance is shared by multiple motors, suppressing the whole machine oscillation caused by sudden load changes of a single motor.

6. The multi-modal cooperative frequency conversion control method for long-distance complex working conditions belt conveyors according to claim 4, characterized in that, While executing the S3 multi-machine asymmetric power balance control step, the S4 energy-saving cruise step based on flow prediction is executed in parallel. Radar level gauges and AI vision acquisition devices are deployed at each material transfer point of the belt conveyor. The radar level gauges and AI vision acquisition devices synchronously collect incoming material information and predict the material flow rate entering the belt within the next T seconds based on the incoming material information. The current actual belt speed of the belt conveyor is read, and the optimal economic belt speed of the belt is calculated in combination with the material flow prediction results. The generalized predictive control algorithm is called, and the optimal economic belt speed is used as the control target. The output frequency of each frequency converter is modified in advance so that the actual belt speed smoothly follows the optimal economic belt speed, reducing the frequent acceleration and deceleration actions caused by sudden changes in material flow.

7. The multi-modal cooperative frequency conversion control method for long-distance complex working conditions belt conveyors according to claim 6, characterized in that, The S4 energy-saving cruise step based on flow prediction establishes the belt conveyor dynamic equation. The state vector of the dynamic equation includes the belt running speed and the belt internal tension, and the control vector of the dynamic equation includes the inverter output torque. In each control cycle, the weighing data collected by the on-site weighing sensor is read, and the material distribution of the entire belt in the next 10 seconds is predicted based on the weighing data. The belt slippage prevention and belt overflow prevention are set as constraint boundaries, and the quadratic programming problem is solved to obtain the optimal motor torque sequence. The control quantity corresponding to the optimal motor torque sequence is sent to each inverter to complete the speed regulation.

8. The multi-modal cooperative frequency conversion control method for long-distance complex working conditions belt conveyors according to claim 7, characterized in that, Upon receiving the conveyor belt stop command, the S5 multi-target coordinated braking step is executed. The S5 multi-target coordinated braking step adopts a composite braking mode combining electrical braking and mechanical braking. Electrical braking is prioritized, and each frequency converter performs DC braking and energy consumption braking. The frequency converter switches to the power generation mode to feed the potential energy of the conveyor belt back to the grid, while simultaneously outputting braking torque. The anti-slip condition is checked in conjunction with the belt slope parameter, and the mechanical kinetic energy of the whole machine is consumed under the premise of meeting the anti-slip condition. The drive motor speed and frequency converter output current are continuously collected. When the drive motor speed drops to the preset critical speed and the frequency converter output current is lower than the preset safe current threshold, the main controller outputs a command to trigger the hydraulic braking system hydraulic brake to complete the mechanical braking intervention.

9. The multi-modal cooperative frequency conversion control method for long-distance complex working conditions belt conveyors according to claim 8, characterized in that, During the operation of the whole machine, if the temperature sensor of a certain drive motor collects an abnormal signal, the main controller will block the collection signal of the abnormal sensor, retrieve the heat conduction simulation data of the motor at the same location obtained from the S1 digital twin prior model construction step to replace the abnormal collection signal, and at the same time issue a frequency reduction operation command to each frequency converter, so that the belt conveyor can maintain continuous conveying operation and achieve non-stop maintenance.

10. A multi-modal cooperative frequency conversion control system for long-distance, complex-condition belt conveyors, used to implement the multi-modal cooperative frequency conversion control method for long-distance, complex-condition belt conveyors as described in any one of claims 1-9, characterized in that, The system includes a main controller unit, a variable frequency drive unit, a drive motor unit, a hydraulic braking unit, and a multi-source sensor acquisition unit. The main controller unit employs a high-performance redundant PLC and internally loads a digital twin program module, a fuzzy inference calculation module, a deviation coupling control calculation module, a generalized predictive control calculation module, a quadratic programming solution module, and a braking logic scheduling module. The variable frequency drive unit contains multiple four-quadrant frequency converters, each electrically connected to a permanent magnet synchronous motor within the drive motor unit. The hydraulic braking unit is equipped with multiple sets of hydraulic disc brakes, with the controlled end of each disc brake connected to the output signal line of the main controller unit. The multi-source sensor acquisition unit... It includes a weighing sensor, a radar level gauge, an AI vision acquisition component, a distributed fiber optic temperature measurement device, a vibration monitor, a speed detection element, a torque detection element, and an ambient temperature detection element. The weighing sensor is installed on the frame of the conveyor belt carrying section. The radar level gauge and the AI ​​vision acquisition component are installed at each material transfer point. The distributed fiber optic temperature measurement device is laid along the entire conveyor belt. The vibration monitor is fixedly installed on the surface of the drive motor and the reducer housing. The speed detection element and the torque detection element are respectively assembled on the output shaft of each drive motor. The ambient temperature detection element is arranged at the conveyor belt operation site. All signal output terminals of the multi-source sensor acquisition unit are connected to the signal input port of the main controller unit.

Images