Belt conveyor self-adaptive soft start control method and system based on load prediction

Abstract

The application provides a kind of adaptive soft start control method and system based on load prediction, before the start of belt conveyor, the initial load weight value on belt conveyor is obtained;Obtain the initial soft start parameter set corresponding to the load interval where the initial load weight value is located, the initial soft start parameter set includes start time, start curve slope and torque limit value;In response to the input of initial soft start parameter set, start the motor of belt conveyor;Within the first preset time period after the start of belt conveyor, real-time collection of actual load weight value and motor current value in operation;Based on the actual load weight value, obtain the fluctuation range of motor current value;Based on the relationship between the collected motor current value and the fluctuation range, correct the initial soft start parameter set;Real-time feedback of the corrected parameters to motor controller;The application effectively prevents belt slip, belt breakage and other faults, prolongs the service life of the equipment, improves the operation efficiency and safety of the whole conveying system.

Description

Technical Field

[0001] This invention relates to the field of belt conveyor technology, and specifically to an adaptive soft-start control method and system for belt conveyors based on load prediction. Background Technology

[0002] High-power, long-distance belt conveyors are core continuous conveying equipment in heavy industries such as coal mining, mining, and ports. They operate in harsh environments with complex and variable load conditions. The start-up phase of the belt conveyor is the most critical and vulnerable link in the entire operation process. The belt conveyor overcomes huge static friction and inertia from a stationary state to smoothly accelerate the unevenly distributed material along the line to the rated speed. It is extremely easy to cause mechanical impact, belt slippage, belt breakage, motor overcurrent and other failures, which seriously affect production safety and efficiency.

[0003] In terms of soft start control, traditional technology generally uses hydraulic couplers to mitigate the current and mechanical shock caused by direct start. Hydraulic couplers transmit torque by adjusting the amount of liquid in the working chamber. Although they have a certain buffering effect, their transmission efficiency is low, making it difficult to achieve precise torque balance under multi-motor drive conditions. Furthermore, they suffer from problems such as oil leakage and high maintenance workload.

[0004] Existing technologies also employ inverter soft-start technology. This technology controls the motor to accelerate along a preset curve by adjusting the output frequency and voltage, significantly reducing the impact on the power grid and mechanical structure. However, the control parameters of this starting method (such as starting time, acceleration curve slope, torque limit value, etc.) are usually preset fixed values, which cannot adapt to the dynamic fluctuations of the load in actual operation. For example, when the actual load is much lower than the preset parameters, the starting process is too slow, resulting in low efficiency and energy waste; while when the actual load is much higher than the preset parameters, insufficient starting torque may cause belt slippage, motor stalling or even burnout, or excessive mechanical impact may damage components such as rollers and idlers, and in severe cases, cause belt breakage accidents, posing a serious threat to safe production.

[0005] In summary, while existing soft-start technologies have alleviated the starting shock problem to some extent, their fixed parameter settings limit the control effectiveness of belt conveyors under complex operating conditions due to dynamic load changes. Therefore, it is necessary to research a load-predictive-based adaptive soft-start control method and system for belt conveyors. Summary of the Invention

[0006] Therefore, the purpose of this invention is to provide an adaptive soft-start control method and system for belt conveyors based on load prediction, which effectively solves the problem that existing high-power, long-distance belt conveyors cannot cope with dynamic load changes during the start-up process, resulting in belt conveyor impact, slippage and overcurrent.

[0007] To achieve the above objectives, the technical solution adopted by this invention is: an adaptive soft-start control method for belt conveyors based on load prediction, comprising:

[0008] Before starting the belt conveyor, obtain the initial load weight value on the belt conveyor;

[0009] In response to the initial load weight value, an initial soft start parameter set corresponding to the load range in which the initial load weight value is located is obtained. The initial soft start parameter set includes the start time, the slope of the start curve, and the torque limit value.

[0010] In response to the input of the initial soft-start parameter set, the motor of the belt conveyor is started;

[0011] During the first preset time period after the belt conveyor starts, the actual load weight value and motor current value are collected in real time during operation.

[0012] Based on the actual load weight value, obtain the fluctuation range of the motor current value;

[0013] Based on the relationship between the collected motor current value and the fluctuation range, the initial soft start parameter set is corrected; the corrected parameters are fed back to the motor controller in real time.

[0014] Furthermore, based on the relationship between the collected motor current values ​​and the fluctuation range, the correction logic for the initial soft-start parameter set is as follows:

[0015] If the motor current value is within the fluctuation range, the motor will continue to be controlled to work according to the input of the initial soft start parameter set;

[0016] If the motor current value is lower than the lower limit of the fluctuation range, slippage detection is performed. If slippage occurs, the slope of the starting curve is reduced and the belt tension is increased. If there is no slippage, the slope of the starting curve is increased and the starting time is shortened.

[0017] If the motor current value is higher than the upper limit of the fluctuation range, the slope of the starting curve will be reduced, the upper limit of torque will be increased, and the starting time will be extended.

[0018] Furthermore, if the current value is higher than the fluctuation range, the motor current value and load weight value are collected within a second preset time. If the motor current value is still higher than the fluctuation range determined by the load weight value, the curve slope is reduced again and an alarm is issued. If the motor current value returns to the fluctuation range, the input is restored to the initial soft start parameter set after the second preset time ends, and the motor continues to be controlled.

[0019] Furthermore, during slippage detection, the motor speed and belt speed are acquired, and a slippage coefficient is obtained based on the motor speed and belt speed. When the slippage coefficient continuously exceeds the safety threshold, the slippage coefficient is continuously collected and judged within a third preset time period. If the slippage coefficient exceeds the safety threshold, the output is a slippage state.

[0020] Furthermore, the starting curve in the initial soft-start parameter set is an S-shaped acceleration curve, and its rate of change of acceleration is limited within a preset range to achieve smooth starting.

[0021] Furthermore, the load values ​​of multiple weighing idlers set on the belt conveyor are collected. Based on the position of the weighing idlers, the corresponding load values ​​are assigned weights, and the load weight value is obtained after weighted fusion.

[0022] Furthermore, a load prediction model is constructed based on historical data, including load weight, motor current, and belt speed. The load weight is predicted using the load prediction model, and the initial soft-start parameter set is planned.

[0023] Furthermore, the initial load weight value is obtained, and a preset load parameter mapping database is called to obtain the set of initial soft start parameters corresponding to the interval of the initial load weight value. The load interval is divided into no-load interval, light load interval, normal load interval and heavy load interval. The initial soft start parameters corresponding to each interval are determined by calibration through historical test data.

[0024] The adaptive soft-start control system for belt conveyors based on load prediction, applied to the above-mentioned methods, includes:

[0025] The load information acquisition unit is used to acquire the initial load weight value on the belt before the belt conveyor starts, and to collect the actual load weight value in real time after the belt conveyor starts.

[0026] The soft start parameter matching unit is connected to the load information acquisition unit and is used to retrieve the corresponding set of initial soft start parameters according to the load range in which the initial load weight value is located.

[0027] A motor starting control unit, connected to the soft start parameter matching unit, is used to start the motor of the belt conveyor according to the initial soft start parameter set;

[0028] The operating status acquisition unit is used to acquire the motor current value in real time;

[0029] The parameter dynamic correction unit is connected to the load information acquisition unit and the operating status acquisition unit. It is used to calculate the fluctuation range of the motor current value based on the actual load weight value, and to correct the initial soft start parameter set according to the relationship between the motor current value and the fluctuation range.

[0030] The feedback unit, connected to the parameter dynamic correction unit and the motor start control unit, is used to feed back the corrected soft start parameters to the motor start control unit in real time to adjust the motor's operating state.

[0031] Furthermore, it also includes a load prediction unit, which is used to predict the load weight value based on historical data and real-time acquired data, and to correct the initial soft start parameter set according to the prediction results.

[0032] The beneficial effects of the above technical solution are as follows: Traditional belt conveyors, due to their fixed control parameters, cannot adapt to dynamic load fluctuations in actual production (caused by factors such as uneven feeding, differences in material properties, and material accumulation), easily leading to problems like impact, slippage, and overcurrent. This invention senses load changes and configures an initial baseline starting rule for the belt conveyor. During startup, it combines the dynamic load changes with the relationship between the motor current and its fluctuation range to configure corresponding control strategies, ensuring that the belt conveyor parameters are always dynamically matched to the actual load conditions. This fundamentally avoids the drawbacks of traditional technologies such as inefficiency under light loads and slippage / overcurrent under heavy loads.

[0033] In its specific implementation, this invention collects load signals by deploying multiple sets of weighing idlers along the conveyor belt's load-bearing section, and accurately calculates the initial load weight value using a weighted fusion algorithm. Subsequently, it combines this with a pre-set load parameter mapping database to retrieve the set of initial soft-start parameters calibrated through historical tests within the corresponding load range. During the start-up execution phase, the motor start-up control unit drives the motor to start according to the initial parameters and an S-shaped acceleration curve, ensuring a smooth application of torque in the initial start-up phase, effectively avoiding mechanical shock, and enabling the targeted setting of start-up strategies based on the initial load conditions.

[0034] During startup, this invention simultaneously collects the actual load weight and motor current values ​​within a first preset time period. Based on the physical correlation between the real-time load and motor torque, it locates the dynamic current fluctuation range. By comparing the deviation between the real-time current and the fluctuation range, it executes a differentiated parameter correction strategy to achieve dynamic control of the belt conveyor from startup to rated speed. The control strategy executes corresponding control measures based on the relationship between current and fluctuation range, matching the corresponding operating conditions, effectively solving problems such as heavy-load slippage, motor stall / burnout, belt breakage, and mechanical impact damage that are prone to occur under traditional fixed-parameter control modes.

[0035] In summary, the adaptive soft-start control method and system for belt conveyors based on load prediction proposed in this invention is a breakthrough optimization of traditional soft-start technology for belt conveyors. It effectively solves the contradiction between fixed parameters and dynamic loads, ensuring the starting safety and stability of high-power, long-distance belt conveyors under complex working conditions, while significantly improving starting efficiency, reducing energy consumption and equipment maintenance costs, and further enhancing the forward-looking nature of the control strategy. It can be widely adapted to belt conveyor application scenarios in heavy industries such as coal mining, mining, and ports, and has significant engineering application value and promotion prospects. Attached Figure Description

[0036] Figure 1 This is a flowchart of the method of the present invention;

[0037] Figure 2 This is a flowchart illustrating the implementation process of load detection.

[0038] Figure 3 This is a flowchart illustrating the implementation process when the motor current exceeds the fluctuation range. Detailed Implementation

[0039] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments:

[0040] Example 1: This example aims to provide an adaptive soft-start control method for belt conveyors based on load prediction, primarily for high-power, long-distance belt conveyors. Traditional soft-start control methods for belt conveyors, such as hydraulic coupling soft start and frequency converter soft start, while mitigating the significant current surge and mechanical stress caused by direct start to some extent, typically use pre-set fixed values ​​for their control parameters (such as start time, start curve slope, and torque limit). However, in actual production, the material load carried by the belt conveyor is dynamically changing, influenced by factors such as uneven feed rate, differences in material characteristics, and material accumulation along the conveyor line. Fixed soft-start parameters cannot adapt to such complex load fluctuations. When the actual load is much lower than the preset parameters, the start-up process becomes too smooth and inefficient; conversely, when the actual load is much higher than the preset parameters, insufficient starting torque may cause belt slippage, motor stalling, or even burnout, or excessive mechanical impact may damage mechanical components such as rollers and idlers, potentially leading to belt breakage accidents and causing significant economic losses and safety risks. Therefore, this embodiment provides an adaptive soft-start control method based on load prediction. By sensing and predicting the load in real time, the starting strategy is dynamically adjusted to fundamentally solve the above problems and achieve safe, efficient, and energy-saving smooth starting.

[0041] When implementing Figure 1As shown in the figure, the belt conveyor adaptive soft start control method based on load prediction in this embodiment abandons the traditional fixed parameter mode. By acquiring the actual load information of the belt conveyor in real time before and at the beginning of the start-up, and predicting the load change trend in the short term based on this, the soft start parameter set (including start-up time, S-curve slope, torque limit value) is dynamically and adaptively adjusted so that the belt start-up process always maintains the best match with the current actual load condition.

[0042] like Figure 2 As shown, load detection is carried out throughout the starting process of the belt conveyor, and the static initial load weight value and the dynamic starting process load weight value are obtained. The initial load weight value is used to obtain the initial soft start parameter set, and the starting process load weight value is used to obtain the fluctuation range of the motor current value during operation, so as to determine the dynamic working conditions during the starting process.

[0043] In specific implementation, such as Figure 2 As shown, before the belt conveyor starts, i.e., in a static environment, the initial load weight value on the belt conveyor is obtained; the load values ​​of multiple sets of weighing idlers installed on the belt conveyor are collected, and based on the position of the weighing idlers, corresponding load values ​​are assigned weights. After weighted fusion, the load weight value is obtained. In implementation, the weighing sensors on each weighing idler convert the detected belt and material pressure into electrical signals. After the signals are amplified by an instrumentation amplifier, they are converted into digital signals by an analog-to-digital converter (ADC) and transmitted to the central processing unit.

[0044] The central processing unit executes a weighted fusion algorithm to calculate the initial load weight value. The specific calculation formula is as follows:

[0045] In the formula, As weight, The signal is the electrical signal acquired for the i-th weighing idler roller.

[0046] In this embodiment, the material distribution on the belt conveyor is often uneven, and the measurement value of a single sensor is easily affected by local fluctuations, potentially resulting in significant errors. This embodiment addresses this by deploying multiple sets of weighing idlers along the conveyor's load-bearing section, forming a distributed sensor array. The weighing sensor on each idler is responsible for collecting the local load value at its location. By weighted fusion of the information collected from each weighing idler, the interference of local fluctuations and single sensor errors on the overall measurement is effectively suppressed, making the initial load weight determination more accurate and laying a solid foundation for subsequent precise control. This is the initial load weight value used by the system and is sent to the adaptive soft-start control algorithm module. In terms of weight configuration, this embodiment adopts the following rules: sensors closer to the loading point are assigned higher weights; sensors with stable signals and low failure rates are assigned lower weights.

[0047] In response to the initial load weight value, the initial soft-start parameter set corresponding to the load range in which the initial load weight value is located is obtained. The specific implementation process is as follows: Figure 2 As shown, the initial load weight value is obtained. Then, it is compared with a preset load threshold range to determine the current starting condition of the belt conveyor. In the starting condition determination, after obtaining the initial load weight value, this embodiment obtains the initial soft start parameter set corresponding to the range of the initial load weight value by calling a preset load parameter mapping database. The load range includes no-load range, light-load range, normal load range and heavy-load range.

[0048] During implementation, the initial soft-start parameters for each interval are determined through calibration using historical test data. Each load interval corresponds to a set of initial soft-start parameters optimized and calibrated using a large amount of historical test data. Based on the accurate load interval determination, the corresponding starting parameters are invoked, enabling the belt conveyor to start according to dynamic rules based on the load conditions.

[0049] The initial soft-start parameter set includes start-up time, start-up curve slope, and torque limit value; in specific implementation, the start-up curve in the initial soft-start parameter set is an S-shaped acceleration curve. In this embodiment, based on the start-up stage of the belt conveyor, the acceleration starts from zero and gradually increases according to a preset smooth curve. In the middle and late stages of the acceleration process, the acceleration smoothly decreases to zero, so that the belt conveyor motor speed reaches the rated speed.

[0050] The rate of change of acceleration is the core parameter of the S-curve and directly affects the smoothness of the start-up. The smaller the rate of change, the gentler the acceleration change, the smaller the impact, and the slightly longer the start-up time.

[0051] The start-up time is the result of the combined effect of load inertia and the slope of the acceleration curve, providing the system with sufficient acceleration time to ensure that the belt conveyor can smoothly reach the rated speed and avoid shock caused by excessively shortening the time in pursuit of efficiency.

[0052] In response to the input of the initial soft start parameter set, the motor of the belt conveyor is started; the initial soft start parameter set configures the starting strategy according to the real-time load condition, so that the belt conveyor motor can start according to the actual condition, ensuring its smooth, efficient and safe start.

[0053] From the start-up to the stable phase of the belt conveyor, within the first preset time period after the belt conveyor starts, such as Figure 2As shown, the actual load weight and motor current values ​​are collected in real time during operation. The actual load weight value is obtained through an array of weighing sensors arranged along the line to obtain real-time load information on the belt conveyor. The motor current value is obtained from the motor drive end through a current sensor (such as a Hall sensor) to directly reflect the motor's output.

[0054] In this embodiment, the expected fluctuation range of the motor current is not determined based on a fixed threshold, but rather on a benchmark value that is dynamically adjusted according to the real-time load. The core principle is that during the soft start of the belt conveyor, the driving torque output by the motor needs to balance the load resistance and the system's acceleration requirements in real time, and this torque has a direct proportional relationship with the motor stator current. Therefore, based on the real-time collected load weight value, combined with factors including the friction coefficient, transmission efficiency, and acceleration, a reasonable fluctuation range of the motor current under the current operating conditions can be obtained.

[0055] Based on the actual load weight, the fluctuation range of the motor current value is obtained. Based on the relationship between the collected motor current value and the fluctuation range, the initial soft-start parameter set is corrected, and the corrected parameters are fed back to the motor controller in real time. In this embodiment, the dynamic expected fluctuation range of the motor current is calculated based on the real-time load weight value, and this is used as a benchmark to compare with the actually collected motor current. If the current value is within this fluctuation range, it indicates that the current soft-start parameters are well matched with the load, and the system maintains the original parameter operation; if the current value is outside the fluctuation range, an anomaly exists, and correction logic is initiated. The specific correction logic is as follows:

[0056] 1. The motor current value is within the fluctuation range.

[0057] If the motor current value is within the fluctuation range, the motor will continue to be controlled according to the input of the initial soft-start parameter set. When the real-time current is within the fluctuation range, it indicates that the current soft-start parameters are well matched with the load conditions, and the motor output torque can smoothly drive the system. At this time, the parameter dynamic correction unit does not intervene, and the motor start control unit continues to execute according to the initial parameter set until the start-up process is completed or the operating conditions change significantly.

[0058] 2. The motor current value is lower than the lower limit of the fluctuation range.

[0059] If the motor current value is lower than the lower limit of the fluctuation range, it indicates that the motor output torque is higher than the actual demand, which may indicate energy waste or slippage risk. The system first performs slippage detection: it acquires the motor speed and belt linear speed in real time and calculates the slippage coefficient (defined as the ratio of the difference between the motor and belt linear speeds to the motor linear speed). If this coefficient continues to exceed the safety threshold within a third preset time period, it is determined to be a slippage state.

[0060] Once slippage is confirmed, the system performs coordinated adjustments: on the one hand, it reduces the slope of the starting curve, and on the other hand, it increases the belt tension through a hydraulic or counterweight tensioning system to increase transmission friction.

[0061] If slippage detection is not triggered, the system is determined to be in a light load state. The system will increase the slope of the start-up curve and shorten the start-up time accordingly to optimize start-up efficiency.

[0062] 3. The motor current value is higher than the upper limit of the fluctuation range.

[0063] like Figure 3 As shown, if the motor current value is higher than the fluctuation range, it indicates that the motor output torque is insufficient to overcome the current load resistance and system inertia of the belt conveyor, and the motor is in an overload operation state. If not intervened in time, it can easily lead to serious faults such as motor overcurrent protection tripping, impact damage to the mechanical structure of the belt conveyor (such as deformation of drums and idlers), or even belt breakage. The essence of the current being higher than the upper limit of the fluctuation range is that the motor output torque is not matched with the actual load demand. Therefore, the slope of the starting curve should be reduced, the upper limit of torque should be increased, and the starting time should be extended.

[0064] Specifically, if the motor current value exceeds the aforementioned fluctuation range, a first-level adjustment strategy is immediately triggered. This strategy involves reducing the slope of the starting curve to minimize dynamic impact, simultaneously increasing the torque limit value, and extending the starting time. In practice, a proportional derating algorithm is used to reduce the slope of the starting curve and minimize dynamic impact. Torque adjustment employs a stepped increase strategy to raise the torque limit value, freeing up the motor's output capacity and ensuring sufficient torque to overcome load resistance. Initially, the torque limit value is raised to 1.1-1.15 times the current value; the upper limit of the torque does not exceed 1.5 times the motor's rated torque to prevent the motor from burning out due to prolonged overload.

[0065] If the current value is higher than the fluctuation range, the motor current value and load weight value are collected within a second preset time. If the motor current value returns to the fluctuation range, the motor is controlled to continue working according to the input of the initial soft start parameter set after the second preset time ends.

[0066] After the primary adjustment strategy is corrected, this embodiment immediately initiates monitoring with a second preset time as the sliding time window. Within the time window, the motor current value and the actual load weight value are periodically collected, and a stability judgment is performed. If, within the second preset time, the real-time current gradually falls back to the fluctuation range, and the actual load weight value... No significant increase indicates that the primary parameter correction is effective and the system load and motor torque have been rematched. After the second preset time, the system starting parameter smoothing reset program gradually restores the corrected slope, torque limit value, and starting time to the set values ​​of the initial soft start parameter set. After the reset is completed, the motor starting control unit continues to operate according to the initial soft start parameter set until the soft start process ends.

[0067] If the motor current value is still higher than the fluctuation range determined by the load weight value, the curve slope will be reduced again and an alarm will be issued. If, at the second preset time interval, the real-time current is still higher than the upper limit of the fluctuation range and the actual load weight value continues to rise, it indicates that the current load has far exceeded the system's first-level adjustment and correction fault tolerance capability, and there is a serious overload risk. In this case, the second-level adjustment strategy will be implemented, and the starting curve slope will be reduced again. Specifically, the starting curve slope will be further reduced to 0.6-0.7 times the current value to minimize acceleration impact. The torque limit value will be maintained at the upper limit after the first-level correction, and the starting time will be extended to 1.3-1.5 times the time after the first-level correction to avoid excessive torque increase leading to motor overheating.

[0068] Based on the degree of current overload, the system triggers a graded early warning mechanism to achieve progressive handling of early warning, intervention, and protection. If it is a minor overload warning, an audible and visual warning signal is triggered, and an overload prompt message is sent to the host computer, prompting maintenance personnel to check the material discharge rate and belt operation status. If it is a moderate overload intervention, in addition to the early warning, the current limiting command of the material discharge device is automatically triggered to reduce the material conveying rate and reduce the load pressure. If it is a severe overload protection, the motor soft stop program is immediately triggered to prevent motor burnout and belt breakage, and an emergency stop alarm message is sent to the host computer.

[0069] This embodiment provides an adaptive soft-start control method based on load prediction for high-power, long-distance belt conveyors. It aims to fundamentally solve the problems of slippage and overcurrent caused by the inability of traditional fixed-parameter soft-start methods to adapt to dynamic load fluctuations, and achieve safe, efficient, and energy-saving smooth start-up.

[0070] Before starting, the system collects data through multiple sets of deployed weighing idlers and calculates the initial load weight value using a weighted fusion algorithm. Then, based on the load range in which the value is located, it matches the corresponding initial soft start parameter set calibrated by historical test data, thereby achieving precise starting under different working conditions.

[0071] After startup, within a preset first time period, the system continuously collects the actual load weight value and motor current value in real time. Based on the actual load weight value, it matches the expected fluctuation range of the motor current. By comparing the real-time collected current value with this dynamic range, it executes differentiated parameter correction logic: if the current value is within the expected range, the existing parameters are maintained; if the current is below the lower limit of the fluctuation range, slippage detection is performed first, and corresponding strategies are executed according to the detection results (slippage or no slippage); if the current is above the upper limit of the fluctuation range, a graded adjustment mechanism is triggered, including first-level real-time correction and continuous second-level monitoring and judgment. If the overload condition persists, the system will further adjust the parameters and trigger graded alarms, thus forming a progressive protection strategy.

[0072] This embodiment ensures that the belt conveyor can always match the actual load from the static start-up state to the dynamic rated speed through real-time sensing and adjustment, which significantly improves the system's adaptability, safety and energy efficiency.

[0073] Example 2

[0074] This embodiment, based on the control method in Embodiment 1, provides an adaptive soft-start control system for belt conveyors based on load prediction. The system includes: a load information acquisition unit, used to acquire the initial load weight value on the belt before the belt conveyor starts, and to collect the actual load weight value during operation in real time after the belt conveyor starts; a soft-start parameter matching unit, connected to the load information acquisition unit, used to retrieve the corresponding initial soft-start parameter set according to the load range where the initial load weight value is located; a motor start control unit, connected to the soft-start parameter matching unit, used to start the motor of the belt conveyor according to the initial soft-start parameter set; an operating status acquisition unit, used to collect the motor current value in real time; a parameter dynamic correction unit, connected to the load information acquisition unit and the operating status acquisition unit, used to calculate the fluctuation range of the motor current value based on the actual load weight value, and to correct the initial soft-start parameter set according to the relationship between the motor current value and the fluctuation range; and a feedback unit, connected to the parameter dynamic correction unit and the motor start control unit, used to feed back the corrected soft-start parameters to the motor start control unit in real time to adjust the operating state of the motor.

[0075] In this embodiment, the various units interact via an industrial bus, forming a complete control closed loop. Before startup, the system matches initial parameters through precise load detection. During startup, the parameters are dynamically corrected based on real-time load and current monitoring. Furthermore, it can be combined with a load prediction unit to achieve forward-looking judgment of load trends and parameter pre-optimization. This significantly improves the accuracy of parameter adaptation and the forward-looking nature of system operation while ensuring smooth and safe startup.

[0076] Example 3

[0077] In this embodiment, based on embodiments 1-2, a load prediction model is constructed based on historical data, including load weight value, motor current value and belt speed value. The load prediction model is used to predict the load weight value and plan the initial soft start parameter set.

[0078] During model training, raw data such as load weight, motor current, and belt speed are collected, and missing values ​​are filled in using interpolation. Multi-dimensional features are then standardized (to eliminate dimensional differences). Subsequently, the data are labeled according to the startup stage, and the parameter matching effect of each stage is correlated to lay the foundation for model training.

[0079] The training model adopts a hybrid architecture of LSTM+XGBoost. The LSTM module is responsible for capturing the time-series dynamic features of parameters such as load and current (such as the continuous change trend of load during startup), while the XGBoost module integrates static features (such as the inherent properties of the belt conveyor) and enhances the nonlinear fitting ability. The two work together to improve prediction accuracy and generalization ability.

[0080] During training, the load, current, and speed sequence before startup are used as inputs, and the actual load at the initial startup stage is used as the prediction target. The system is divided into training, validation, and test sets. The model performance is evaluated by RMSE (root mean square error) and MAPE (mean absolute percentage error), and an incremental learning strategy is adopted to ensure that the model adapts to changes in operating conditions.

[0081] Based on the prediction results, the system classifies the load trend into three categories:

[0082] Stable type: Small load fluctuation, using reference parameters;

[0083] As the load continues to increase, the starting time should be extended accordingly, the slope of the starting curve should be reduced, and the torque limit should be appropriately increased.

[0084] Decreasing type: As the load gradually decreases, the start-up time is shortened and the slope is increased.

[0085] The planned parameters will be simulated and verified on the digital twin platform. Only after confirming that there are no risks such as slippage or overload will they be sent to the motor controller for execution.

[0086] To accommodate the content of this embodiment, a load prediction unit is added to the system framework, which can continuously optimize the initial soft-start parameter set based on historical and real-time data. This method upgrades the control strategy from traditional static preset and passive response to dynamic prediction and active adaptation, effectively overcoming the bottleneck that fixed parameters cannot adapt to dynamic load changes, and significantly improving the safety, stability, and energy efficiency of the startup process.

[0087] Therefore, this embodiment constructs a load prediction model based on historical operating data and real-time acquired data. This model can predict the load change trend during the period from start-up to rated speed, thereby transforming passive response control into proactive look-ahead control. Through the predicted load curve, key control parameters during the soft start process are dynamically calculated and optimized, such as the motor's target speed curve, torque increase, and acceleration time, ensuring smooth, efficient, and safe start-up under any load conditions. Seamless integration of the load prediction module with the soft start drive device enables control from data acquisition, processing, and prediction to control command generation and execution. This reduces reliance on operator experience, improves control accuracy and reliability, significantly reduces mechanical shock and electrical stress during conveyor belt start-up, effectively prevents belt slippage and breakage, extends equipment lifespan, improves the overall efficiency and safety of the conveyor system, and simultaneously reduces energy consumption and maintenance costs.

Claims

1. A method for adaptive soft-start control of belt conveyors based on load prediction, characterized in that: include Before starting the belt conveyor, obtain the initial load weight value on the belt conveyor; In response to the initial load weight value, an initial soft start parameter set corresponding to the load range in which the initial load weight value is located is obtained. The initial soft start parameter set includes the start time, the slope of the start curve, and the torque limit value. In response to the input of the initial soft-start parameter set, the motor of the belt conveyor is started; During the first preset time period after the belt conveyor starts, the actual load weight value and motor current value are collected in real time during operation. Based on the actual load weight value, obtain the fluctuation range of the motor current value; Based on the relationship between the collected motor current value and the fluctuation range, the initial soft start parameter set is corrected; the corrected parameters are fed back to the motor controller in real time.

2. The adaptive soft-start control method for belt conveyors based on load prediction according to claim 1, characterized in that: Based on the relationship between the collected motor current values ​​and the fluctuation range, the correction logic for the initial soft-start parameter set is as follows: If the motor current value is within the fluctuation range, the motor will continue to be controlled to work according to the input of the initial soft start parameter set; If the motor current value is lower than the lower limit of the fluctuation range, slippage detection is performed. If slippage occurs, the slope of the starting curve is reduced and the belt tension is increased. If there is no slippage, the slope of the starting curve is increased and the starting time is shortened. If the motor current value is higher than the upper limit of the fluctuation range, the slope of the starting curve will be reduced, the upper limit of torque will be increased, and the starting time will be extended.

3. The adaptive soft-start control method for belt conveyors based on load prediction according to claim 2, characterized in that: If the current value is higher than the upper limit of the fluctuation range, the motor current value and load weight value are collected within a second preset time. If the motor current value is still higher than the fluctuation range determined by the load weight value, the curve slope is reduced again and an alarm is issued. If the motor current value returns to the fluctuation range, after the second preset time ends, the input will be restored to the initial soft start parameter set, and the motor will continue to be controlled.

4. The adaptive soft-start control method for belt conveyors based on load prediction according to claim 2, characterized in that: During slippage detection, the motor speed and belt speed are acquired, and the slippage coefficient is obtained based on the motor speed and belt speed. When the slippage coefficient continuously exceeds the safety threshold, the slippage coefficient is continuously collected and judged within a third preset time period. If the slippage coefficient exceeds the safety threshold, the output is a slippage state.

5. The adaptive soft-start control method for belt conveyors based on load prediction according to claim 1, characterized in that: The starting curve in the initial soft-start parameter set is an S-shaped acceleration curve, and its rate of change of acceleration is limited within a preset range to achieve smooth start-up.

6. The adaptive soft-start control method for belt conveyors based on load prediction according to claim 1, characterized in that: The load values ​​of multiple weighing idlers set on the belt conveyor are collected. Based on the position of the weighing idlers, the corresponding load values ​​are assigned weights, and the load weight value is obtained after weighted fusion.

7. The adaptive soft-start control method for belt conveyors based on load prediction according to claim 1, characterized in that: A load prediction model is constructed based on historical data, including load weight, motor current, and belt speed. The load weight is predicted using the load prediction model, and the initial soft start parameter set is planned.

8. The adaptive soft-start control method for belt conveyors based on load prediction according to claim 1, characterized in that: The initial load weight value is obtained, and the preset load parameter mapping database is called to obtain the set of initial soft start parameters corresponding to the range of the initial load weight value. The load range is divided into no-load range, light load range, normal load range and heavy load range. The initial soft start parameters corresponding to each range are determined by calibration through historical test data.

9. A belt conveyor adaptive soft-start control system based on load prediction, applied to the method described in any one of claims 1-8, characterized in that, include: The load information acquisition unit is used to acquire the initial load weight value on the belt before the belt conveyor starts, and to collect the actual load weight value in real time after the belt conveyor starts. The soft start parameter matching unit is connected to the load information acquisition unit and is used to retrieve the corresponding set of initial soft start parameters according to the load range in which the initial load weight value is located. A motor starting control unit, connected to the soft start parameter matching unit, is used to start the motor of the belt conveyor according to the initial soft start parameter set; The operating status acquisition unit is used to acquire the motor current value in real time; The parameter dynamic correction unit is connected to the load information acquisition unit and the operating status acquisition unit. It is used to calculate the fluctuation range of the motor current value based on the actual load weight value, and to correct the initial soft start parameter set according to the relationship between the motor current value and the fluctuation range. The feedback unit, connected to the parameter dynamic correction unit and the motor start control unit, is used to feed back the corrected soft start parameters to the motor start control unit in real time to adjust the motor's operating state.

10. The adaptive soft-start control system for belt conveyors based on load prediction according to claim 9, characterized in that: It also includes a load prediction unit, which is used to predict the load weight value based on historical data and real-time acquired data, and to correct the initial soft start parameter set according to the prediction results.

Images