Self-adaptive branching angle control method of high-speed branching machine for roller conveying line

Abstract

The invention relates to the technical field of high-speed branching machine self-adaptive branching control, in particular to a high-speed branching machine self-adaptive branching angle control method for a roller conveying line, and the method comprises the steps: carrying out the differential angle calculation based on the size data of a turnover box and the type of a direction instruction; a right steering angle is generated according to the right shunt instruction, and a zero angle is output according to the straight instruction. And when the front end of the turnover box reaches the trigger position, the servo motor is driven to enable the shunt wheel to rotate to a target angle and maintain a constant elevation. The angle offset is detected in real time, when a photoelectric trigger signal indicating that the turnover box leaves is responded, the shunting wheel resetting and returning-to-zero process and the new turnover box entering process are synchronously executed, a graded safety control strategy is implemented based on a pressure gradient parameter and a three-dimensional position offset monitoring result, a historical database is associated to iteratively optimize model parameters, and the safety of the turnover box is improved. And the branching efficiency and the operation reliability are improved.

Description

Technical Field

[0001] This invention relates to the field of adaptive branching control technology for high-speed branching machines, and more particularly to an adaptive branching angle control method for high-speed branching machines used in roller conveyor lines. Background Technology

[0002] Sorting machines are widely used in factory logistics automation systems, primarily to work with roller conveyors to change the direction of goods transport. In existing technology, sorting machines mainly consist of a cam-lifting motor and a sorting wheel conveyor motor. The operation process is as follows: before the goods arrive at the sorting section, the cam-lifting motor drives the sorting machine to its upper limit position. Subsequently, the sorting wheel conveyor motor drives the sorting wheel to rotate via an O-belt, thereby turning the goods to the designated conveyor line. Common application scenarios include 30-degree left or right turns to transfer goods from one conveyor line to another. This structure relies on a fixed turning angle, and the sequence of actions must strictly follow the lifting and rotating steps, making it the mainstream design in the current logistics conveying field.

[0003] Existing technologies suffer from several pain points in practical use, primarily in terms of low efficiency and insufficient adaptability. First, the fixed steering angle mechanism cannot dynamically adjust according to cargo size, leading to insufficient or redundant steering force and causing jamming issues. For example, when conveying large tote boxes, a fixed 30-degree angle cannot provide sufficient steering torque, causing cargo to easily stall and become congested in the distribution section. Conversely, small tote boxes may over-steer at the same angle, increasing the risk of instability in transport. Second, the control method suffers from slow cycle time because the distribution wheels cannot rotate until the distribution machine is fully raised, causing a delay. For instance, in continuous transport scenarios, this accumulated waiting time significantly reduces overall throughput, failing to meet the demands of high-speed logistics. These problems collectively limit the versatility and efficiency of the distribution equipment, necessitating improvements to enhance automation levels. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides an adaptive branching angle control method for high-speed branching machines in roller conveyor lines, solving the technical problems of adaptive steering and high-speed operation of branching machines.

[0005] To solve the above-mentioned technical problems, the specific technical solution of the present invention is as follows:

[0006] The present invention provides an adaptive branching angle control method for a high-speed branching machine in a roller conveyor line, comprising:

[0007] Obtain the length, width, and conveying direction data of the turnover box;

[0008] Call the branch unit segment accumulation parameters;

[0009] Performing a branch rotation angle calculation based on length data, width data and accumulation parameter to generate a steering angle output value;

[0010] When the front end of the turnover box reaches the branch section trigger position, the control system drives the branch wheel to rotate to the steering angle output value;

[0011] The turnover box is transported to the branch completion state through the friction of the branch wheel surface;

[0012] Real-time detection of the photoelectric trigger signal of the turnover box leaving the branch machine area;

[0013] In response to the photoelectric trigger signal, the branch wheel reset and zero and the new turnover box feeding process are synchronously executed.

[0014] Further, the adaptive branch angle control method for the high-speed branch machine of the drum conveying line comprises:

[0015] Receiving the conveying direction instruction issued by the upper computer as a direction control input;

[0016] Scanning the length characteristics of the turnover box through the first photoelectric array arranged longitudinally along the drum conveying line;

[0017] Capturing the width distribution of the turnover box through the second photoelectric array perpendicular to the conveying direction;

[0018] Performing a box posture scanning process based on the length characteristics and the width distribution;

[0019] When the box posture scanning confirms that the turnover box enters the identification area, the length characteristics and the width distribution are aggregated as a size data set;

[0020] Integrating the size data set and the direction control input into a transmission data packet, and transmitting the transmission data packet to the control execution unit.

[0021] Further, the adaptive branch angle control method for the high-speed branch machine of the drum conveying line comprises:

[0022] Obtaining the branch machine section accumulation parameter as a basic input data, and analyzing the size data set to extract the length value and the width value of the turnover box as calculation parameters;

[0023] Identifying the conveying direction instruction type as a left branch instruction, a right branch instruction or a straight instruction, and performing a direction adaptive calculation process based on the identification result:

[0024] When the left branch instruction is identified, the calculation parameters are input into a length-dominated calculation process to generate a left steering angle;

[0025] When identified as a right branch instruction, the calculation parameter input width optimization type calculation process generates a right turning angle;

[0026] When identified as a straight instruction, skip the calculation process and directly output a zero angle instruction;

[0027] Output the steering angle control instruction corresponding to the direction to the rotation control system, and update the calculation rule adaptation parameter database based on the calculation parameter.

[0028] Further, the high-speed branch angle control method for the high-speed branch machine of the drum conveying line of the application, when the front end of the turnover box reaches the branch segment trigger position, the control system drives the branch wheel to rotate to the turning angle output value, including:

[0029] Real-time capture the displacement change parameter of the conveying line as displacement monitoring data, input the displacement monitoring data into the position prediction algorithm to generate a position prediction trajectory;

[0030] Input the position prediction trajectory into the posture compensation mechanism to optimize the angle positioning accuracy, and generate real-time posture compensation instructions for the branch wheel according to the optimization results;

[0031] Input the posture compensation instructions into the rotation execution component to control the posture calibration positioning, constantly maintain the branch wheel contact surface elevation parameter during the posture calibration positioning process, collect the rotation positioning dynamic deviation data and return to the posture compensation mechanism optimization process.

[0032] Further, the high-speed branch angle control method for the high-speed branch machine of the drum conveying line of the application, the branch wheel turning angle calculation based on length data, width data and accumulation parameter, generate the turning angle output value, also includes:

[0033] Collect the branch machine segment accumulation parameter and size data as input data, and extract the turnover box length value from the input data in the previous step as the dominant calculation weight factor;

[0034] Input the dominant calculation weight factor into the multi-parameter fusion operation processing to generate the turning angle intermediate value;

[0035] Call the positioning offset correction parameter in the posture compensation mechanism, combine the positioning offset correction parameter with the previous step angle intermediate value to generate the compensated control instruction;

[0036] Store the dominant weight factor distribution trajectory to the historical database, and update the multi-parameter fusion rule adaptation left turning physical feature library based on the distribution trajectory.

[0037] Further, the high-speed branching machine adaptive branching angle control method for the drum conveying line disclosed by the present application, the branching wheel turning angle calculation based on length data, width data and accumulation parameter generates turning angle output value, and further comprises:

[0038] The branching machine segment accumulation parameter and size data set are taken as calculation input data, the turnover box width value is extracted from the input data of the previous step as the core calculation weight factor, and the rotation direction physical constraint model is constructed based on the core calculation weight factor;

[0039] The physical constraint model is input into multi-parameter fusion operation processing to generate a turning angle intermediate value;

[0040] The left turn physical compensation rule in the historical feature library is called, the left turn physical compensation rule is combined with the angle intermediate value of the previous step to generate a corrected control instruction;

[0041] Real-time turning trajectory feature data is captured through a trajectory monitoring device, and the trajectory feature data is compared with a preset tolerance model for verification processing, when the verification result is trajectory deviation, dynamic compensation parameters are generated according to the deviation, the dynamic compensation parameters are stored in a database and a rule iteration optimization instruction is triggered.

[0042] Further, the high-speed branching machine adaptive branching angle control method for the drum conveying line disclosed by the present application, the branching wheel surface friction conveying turnover box comprises:

[0043] The rotational speed state parameter of the branching wheel driving unit is collected in real time as dynamic input data, and a torque compensation instruction is generated according to the fluctuation of the dynamic input data;

[0044] The torque compensation instruction is input into a friction contact surface pressure prediction model, and the friction contact surface pressure distribution change trend is output through the pressure prediction model;

[0045] The rotational output torque is dynamically adjusted based on the pressure distribution change trend, and the actual angle deviation of the branching wheel is detected in real time;

[0046] When the actual angle deviation exceeds the safety threshold, the following is performed: lock the current branching wheel rotation trajectory, activate the emergency correction control process, generate the state response instruction of the branching wheel zero action linkage, store the dynamic parameter set of the conveying process and associate the positioning deviation data.

[0047] Further, the high-speed branching machine adaptive branching angle control method for the drum conveying line disclosed by the present application, the response photoelectric trigger signal further comprises:

[0048] The motion posture type of the turnover box leaving process is identified through a motion capture device, and the motion posture type is input into a reset timing classification model to generate a differentiated reset instruction;

[0049] Synchronous scanning shunt section retention detection signal strength parameter, when the retention signal strength parameter exceeds the safety threshold, the auxiliary oscillation cleaning mechanism is started to process the retention, and an emergency channel switching control instruction is generated

[0050] The process parameter of the shunt wheel zero reset action is associated with the historical database record, the preventive maintenance evaluation process of the conveying equipment is triggered based on the associated data, the system event log is updated, and the dynamic parameter package is compressed and stored in the maintenance database.

[0051] Further, the high-speed shunting machine adaptive shunting angle control method for the drum conveying line disclosed by the present application, the synchronous execution of the shunting wheel reset zero and the new turnover box feeding process start in response to the photoelectric trigger signal, further comprises:

[0052] Collect real-time safety state evaluation parameters of the shunting section as risk judgment inputs, load the risk judgment inputs into a risk level judgment process to generate a risk level identifier;

[0053] According to the risk level identifier, perform branch response operations, when the risk level identifier is low risk, generate a conveying unit fast start instruction, activate the pre-stage blocker release sequence, and reconstruct the conveying channel state matrix;

[0054] When the risk level identifier is high risk, start a safety buffer delay mechanism, and generate a channel defect repair instruction set;

[0055] Associate the throughput counter value with the historical maintenance database record, iteratively optimize the risk level judgment threshold parameter based on the associated value, compress and store the state reconstruction process data package, update the maintenance log, and feed back the optimized threshold parameter to the risk level judgment process.

[0056] Further, the high-speed shunting machine adaptive shunting angle control method for the drum conveying line disclosed by the present application, the conveying turnover box to the shunting completion state through the surface friction of the shunting wheel comprises:

[0057] Real-time collection of three-dimensional position offset dynamic parameters of the shunting wheel as spacing control inputs, loading the spacing control inputs into a pressure distribution feedback model to generate pressure gradient parameters;

[0058] According to the pressure gradient parameter, perform safety state judgment, when the pressure gradient parameter is within the safety threshold range, start the high-frequency micro-vibration suppression balance period to stabilize the contact surface, and activate the rolling friction compensation calibration process;

[0059] When the pressure gradient parameter exceeds the safety threshold range, trigger the mechanism rigid locking protection action, and solve the mechanical deformation correction vector parameter;

[0060] The associated equipment operation height parameter is used to generate a calibration benchmark with a historical positioning database record, a pressure distribution feedback model parameter is dynamically reconstructed according to the calibration benchmark, a position offset accumulation trajectory parameter package is generated and compressed and stored into a maintenance database, and the stored parameter is fed back to a pressure distribution feedback model optimization process.

[0061] Advantages of the present application

[0062] The technical solution provided by the present application achieves the following advantages: a calculation model is used to dynamically collect and accumulate parameters based on the size characteristics of the turnover box, a differentiated turning angle calculation process is triggered based on the type of the conveying direction instruction, the output angle of the branch wheel is accurately adapted to the turning mechanical characteristics of the box of different sizes, and the problems of box jamming or excessive turning caused by the fixed angle mechanism are effectively eliminated; through the cooperative control of the integrated servo drive and the position prediction algorithm, direct turning in conveying is realized under the condition of maintaining the constant elevation of the branch wheel, the time delay of the traditional lifting action is eliminated, and the branch beat is significantly compressed; the friction contact surface pressure prediction model and the high-frequency micro-vibration suppression technology are combined to dynamically optimize the rotary torque output and stabilize the contact surface, thereby enhancing the high-speed conveying stability; a closed-loop safety optimization mechanism is constructed, a hierarchical risk response strategy is triggered through pressure gradient monitoring and position offset analysis, the risk judgment threshold and the pressure distribution model parameter are iteratively updated in association with the historical operation data of the equipment, the system reliability and the self-adaptive ability are continuously improved, and finally the branch efficiency and the equipment durability are cooperatively optimized. BRIEF DESCRIPTION OF DRAWINGS

[0063] In order to more clearly illustrate the technical solutions of the present application, the following will briefly introduce the drawings needed to be used in the embodiments. Obviously, other drawings can also be obtained by those skilled in the art without any creative labor under the premise of the drawings.

[0064] Figure 1 The high-speed branch machine electrical scheme connection block diagram for the high-speed branch angle control method of the high-speed branch machine for the drum conveying line provided by the embodiment of the present application.

[0065] Figure 2 The high-speed branch machine flow direction block diagram for the high-speed branch angle control method of the high-speed branch machine for the drum conveying line provided by the embodiment of the present application.

[0066] Figure 3 The high-speed branch machine control method diagram for the high-speed branch angle control method of the high-speed branch machine for the drum conveying line provided by the embodiment of the present application.

[0067] Figure 4 The symbol identification diagram for the high-speed branch angle control method of the high-speed branch machine for the drum conveying line provided by the embodiment of the present application. DETAILED DESCRIPTION

[0068] In order to make the objects, technical solutions and advantages of the present application clearer, the technical solutions of the present application will be described clearly and completely below in combination with specific embodiments of the present application and corresponding drawings. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work belong to the protection scope of the present application. The technical solutions provided by the embodiments of the present application will be described in detail below in combination with the drawings. In order to better understand the objects of the present application, the present application will be further described in detail below.

[0069] Please refer to Figure 1 The present application provides a high-speed shunting machine adaptive shunting angle control method for a drum conveying line, which comprises the following steps of:

[0070] acquiring length data, width data and conveying direction instructions of the turnover box;

[0071] calling shunting machine section accumulation parameters;

[0072] performing shunting wheel turning angle calculation based on the length data, the width data and the accumulation parameters to generate a turning angle output value;

[0073] when the front end of the turnover box reaches a shunting section trigger position, the control system drives the shunting wheel to rotate to the turning angle output value;

[0074] the turnover box is conveyed to a shunting completion state through the surface friction of the shunting wheel;

[0075] real-time detection of a photoelectric trigger signal of the turnover box leaving the shunting machine area;

[0076] synchronous execution of shunting wheel reset zero and new turnover box feeding process starting in response to the photoelectric trigger signal.

[0077] The high-speed shunting machine adaptive shunting angle control method for the drum conveying line provided by the present application firstly involves the acquisition process of the size data of the turnover box and the conveying direction instructions. The process receives the conveying direction instructions issued by the upper computer as the direction control input, scans the length features of the turnover box by using the photoelectric array arranged along the longitudinal direction of the drum conveying line, simultaneously captures the width distribution of the turnover box by using the photoelectric array perpendicular to the conveying direction, performs the box posture scanning processing to confirm the entry of the turnover box into the identification area, aggregates the length features and the width distribution as the size data set, and integrates it into a transmission data packet to be transmitted to the control execution unit, thereby providing basic data support for subsequent calculation. This step realizes accurate collection of size information and timely transmission of instructions, and lays a data foundation for adaptive control.

[0078] Then, the control system calls the accumulation and storage parameters of the shunting machine section, including key physical dimensions such as accumulation and storage width and wheel spacing, as input basis for the calculation process. These parameters are pre-stored in the system database, and the relevant values are obtained by accessing the database for subsequent calculation and processing of the turning angle. The calling of accumulation and storage parameters helps to combine the actual device structure with the size of the turnover box, improve the adaptability and accuracy of the turning calculation, and avoid shunting errors caused by fixed characteristics of the device.

[0079] Based on the length data, width data and accumulation and storage parameters, the system performs shunting wheel turning angle calculation processing. This process analyzes the size data set to extract the length and width values of the turnover box as calculation parameters, identifies the direction instruction type as left shunting instruction, right shunting instruction or straight instruction, and performs direction adaptive calculation processing according to different instruction types: when identifying as left shunting instruction, input length dominant calculation process to generate left turning angle; when identifying as right shunting instruction, input width optimization calculation process to generate right turning angle; when identifying as straight instruction, directly output zero angle instruction. After calculation is completed, the corresponding direction turning angle control instruction is output to the rotation control system, and the calculation rule adaptation parameter database is updated based on the calculation parameters of this time, realizing dynamic optimization and iterative improvement of the algorithm.

[0080] When the front end of the turnover box reaches the triggering position of the shunting section, the control system drives the shunting wheel to rotate to the turning angle output value. This operation captures the displacement change parameter of the conveying line in real time as displacement monitoring data, inputs the position prediction algorithm to generate a position prediction trajectory, inputs the position prediction trajectory into the posture compensation mechanism to optimize the angle positioning accuracy, generates a shunting wheel real-time posture compensation instruction, and inputs the posture compensation instruction into the rotation execution component to control the posture calibration positioning. In the posture calibration positioning process, the shunting wheel contact surface elevation parameter is constantly maintained, the rotation positioning dynamic deviation data is collected and fed back to the posture compensation mechanism optimization process, the accuracy and timeliness of the rotation action are realized, and the shunting delay is reduced.

[0081] Subsequently, the turnover box is conveyed to the shunting completion state through the surface friction of the shunting wheel. This conveying process collects the rotation speed state parameter of the shunting wheel drive unit in real time as dynamic input data, generates a torque compensation instruction according to the fluctuation amount, inputs the torque compensation instruction into the friction contact surface pressure prediction model to output the friction contact surface pressure distribution trend, dynamically adjusts the rotation output torque based on the pressure distribution trend, and detects the actual angle deviation of the shunting wheel in real time. When the actual angle deviation exceeds the safety threshold, the current shunting wheel rotation trajectory is locked, the emergency correction control process is activated, the state response instruction of the shunting wheel zero action linkage is generated, the dynamic parameter set of the conveying process is stored and associated with the positioning deviation data, and the stability and continuity of the conveying are ensured.

[0082] Real-time detection of the photoelectric trigger signal of the turnover box leaving the branching machine area, using a photoelectric sensor to capture the movement state change of the turnover box, generating a photoelectric trigger event input to the control system. This detection process serves as a symbol of branching completion, triggering the subsequent response mechanism and providing a time reference for system beat optimization.

[0083] In response to the photoelectric trigger signal, the system synchronously executes the branching wheel reset and the new turnover box feeding process. This response includes identifying the motion posture type of the turnover box during the departure process through a motion capture device, inputting the reset timing classification model to generate differentiated reset instructions, synchronously scanning the signal strength parameters of the branching segment retention, and starting the auxiliary oscillation cleaning mechanism to process the retention when the signal strength parameters exceed the safety threshold. Generate emergency channel switching control instructions. At the same time, collect real-time safety state evaluation parameters of the branching segment as risk judgment input, load into the risk level judgment process to generate risk level identification, and execute branch response operation according to the risk level identification: when the risk is low, generate a conveying unit fast start instruction to activate the pre-segment blocker to release the sequence to rebuild the conveying channel state matrix; when the risk is high, start the safety buffer delay mechanism to generate a set of channel defect repair instructions. Correlate the process parameters of the branching wheel reset action with the historical database records, trigger the preventive maintenance evaluation process of the conveying equipment, update the system event log and compress the dynamic parameter package to the maintenance database, and realize seamless connection of the branching wheel reset and the new process.

[0084] Specifically, the adaptive branching angle control method for high-speed branching machines of drum conveying lines described in the present application includes:

[0085] Receiving the conveying direction instruction issued by the upper computer as a direction control input;

[0086] Scanning the length characteristics of the turnover box through the first photoelectric array arranged longitudinally along the drum conveying line;

[0087] Capture the width distribution of the turnover box through the second photoelectric array perpendicular to the conveying direction;

[0088] Performing box posture scanning processing based on length characteristics and width distribution;

[0089] When the box posture scanning confirms that the turnover box enters the identification area, aggregate the length characteristics and width distribution into a size data set;

[0090] Integrate the size data set and the direction control input into a transmission data packet, and transmit the transmission data packet to the control execution unit.

[0091] The direction control input is received from the host computer, which generates a command based on the transport path planning, including left branch, right branch or straight operation type. The host computer sends the command to the local controller through an industrial communication protocol such as Modbus or EtherCAT, serving as the initial guidance input for the branch operation. This step provides a reference for the branch direction, avoiding manual intervention and improving automation level.

[0092] The first photoelectric array is arranged longitudinally along the roller conveyor line to scan the length feature of the toter, which is composed of multiple linearly arranged infrared photoelectric sensors installed on both sides of the conveyor line in the transport direction. When the toter passes through, the sensors detect the time difference of the box blocking, and calculate the length of the toter combined with the speed parameter of the conveyor line. This scanning process captures length changes in real time, providing dynamic input for subsequent size aggregation and reducing measurement errors.

[0093] The second photoelectric array is arranged perpendicular to the transport direction to capture the width distribution of the toter, which covers the width range of the conveyor line with an array of light curtain sensors. The array detects changes in the light signal of the toter's lateral profile to analyze the width distribution feature. This capture works in conjunction with the first array to achieve synchronous collection of two-dimensional size and enhance data integrity.

[0094] Based on the length feature and width distribution, the toter posture scanning process is performed, and the control system calls the scanning algorithm to process the photoelectric signal to identify the position offset and rotation angle of the toter. The algorithm includes noise filtering and edge detection modules to eliminate transport vibration interference and confirm the stability of the toter posture. This processing step verifies the reliability of the size data and provides a basis for entering the recognition area.

[0095] When the toter posture scanning confirms that the toter has entered the recognition area, the length feature and width distribution are aggregated into a size data set. The system defines the recognition area as a fixed detection segment, and marks the entry point of the toter with a photoelectric trigger signal. The aggregation module integrates the length and width data into a structured data set, which is stored in JSON or binary format. This set serves as a unified size output, facilitating subsequent transmission and processing.

[0096] The size data set and direction control input are integrated into a transmission data packet, which is transmitted to the control execution unit. The integration process includes data packaging and verification, adding a timestamp and serial number to generate a transmission data packet. It is transmitted to the branch machine control execution unit through industrial Ethernet or CAN bus. This step realizes efficient data transmission and supports real-time control decisions.

[0097] Specifically, the high-speed shunting machine adaptive shunting angle control method for the drum conveying line comprises the following steps of:

[0098] Obtaining shunting machine section accumulation parameter as basic input data, analyzing size data set to extract length value and width value of the turnover box as calculation parameters;

[0099] Identifying the conveying direction instruction type as left shunting instruction, right shunting instruction or straight instruction, and performing direction adaptive calculation processing based on the identification result:

[0100] When the left shunting instruction is identified, the calculation parameters are input into the length-dominant calculation process to generate a left turning angle;

[0101] When the right shunting instruction is identified, the calculation parameters are input into the width-optimized calculation process to generate a right turning angle;

[0102] When the straight instruction is identified, the calculation process is skipped and a zero angle instruction is directly output;

[0103] Outputting the turning angle control instruction corresponding to the direction to the rotation control system, and updating the calculation rule adaptation parameter database based on the calculation parameters.

[0104] The shunting wheel turning angle calculation process comprises the following steps of:

[0105] Analyzing size data set to extract length value and width value of the turnover box as calculation parameters. The size data set comes from the photoelectric scanning processing result, and the analysis module separates the length and width values and converts them into floating point format. This step converts the original size data into calculable variables, eliminates data redundancy, and provides standardized input for direction adaptive calculation.

[0106] Identifying the conveying direction instruction type as left shunting instruction, right shunting instruction or straight instruction. The instruction recognition module analyzes the direction identifier in the data packet transmitted by the upper computer, and the classification processing logic includes instruction validity check and abnormal instruction filtering. The classification provides branch judgment basis for subsequent calculation, and ensures the determinacy of the processing path.

[0107] Perform direction adaptive calculation processing based on the recognition result: when recognized as a left branch instruction, input the length parameter into the length-dominant calculation process to generate a left turning angle. This calculation process takes the length of the turnover box as the main weight factor, combines the width parameter, and outputs the left turning angle value through the inverse trigonometric function model, adapting to the turning mechanical characteristics of long-size boxes.

[0108] When recognized as a right branch instruction, input the width parameter into the width-optimized calculation process to generate a right turning angle. This process prioritizes the width distribution characteristics of the turnover box, constructs a rotation direction physical constraint model, and generates a right turning angle value through multi-parameter fusion operation. The calculation process calls the left turning physical compensation rules of the historical feature library for angle correction to optimize the stability of the turning trajectory of short and wide boxes.

[0109] When recognized as a straight-line instruction, skip the calculation process and directly output a zero-angle instruction. This processing path bypasses the complex calculation link and sets the target angle of the rotation control system to zero degrees as the reference position. The straight-line mode uses a short-circuit logic response to reduce system calculation load and improve real-time control efficiency.

[0110] Output the corresponding direction turning angle control instruction to the rotation control system. The instruction output module encapsulates the calculated angle value into a servo motor control protocol format, including the target angle and rotation direction identifier. Through real-time bus transmission to the branch wheel driving unit, the servo motor is controlled to perform accurate rotation positioning.

[0111] Update the parameter database based on the current calculation parameter update calculation rule. The data update module records the current turnover box length value, width value, and final output angle, and collects actual turning trajectory characteristic data through the trajectory monitoring device. When a trajectory deviation is detected, a dynamic compensation parameter is generated, triggering the rule iteration optimization instruction to update the physical feature library, realizing algorithm adaptive evolution.

[0112] Specifically, the adaptive branch angle control method for the high-speed branch machine of the drum conveying line provided by the present application comprises the following steps:

[0113] Real-time capture the displacement change parameter of the conveying line as displacement monitoring data, and input the displacement monitoring data into the position prediction algorithm to generate a position prediction trajectory;

[0114] Input the position prediction trajectory into the posture compensation mechanism to optimize the angle positioning accuracy, and generate a real-time posture compensation instruction for the branch wheel according to the optimization result;

[0115] Input the posture compensation instruction into the rotation execution component to control the posture calibration positioning, and constantly maintain the contact surface elevation parameter of the branch wheel during the posture calibration positioning process, collect the rotation positioning dynamic deviation data and return it to the posture compensation mechanism optimization process.

[0116] Real-time displacement change parameters of the conveying line are captured as displacement monitoring data, and the process is implemented by an incremental encoder or a laser displacement sensor installed on the side of the conveying line to detect real-time position changes in the movement of the toter. The displacement monitoring data includes a speed vector and an acceleration component, and the system inputs these data into a position prediction algorithm such as a Kalman filter model based on time series to generate a position prediction trajectory. The prediction trajectory describes the time point and spatial coordinates of the toter arriving at the branching section in the future, providing time buffer for the advance adjustment of the branching wheel and reducing response delay.

[0117] The position prediction trajectory is input into the posture compensation mechanism to optimize the angle positioning accuracy. The posture compensation mechanism adopts a closed-loop feedback control strategy, including a proportional-integral-derivative (PID) regulator module. The prediction trajectory is used as an input parameter, and the compensation mechanism combines the current posture angle data of the branching wheel to calculate the angle deviation and generate real-time posture compensation instructions for the branching wheel. The compensation instructions include an angle correction value and a rotation direction identifier, which improve the positioning accuracy of the branching wheel through the optimization process and avoid the steering angle deviation caused by conveying vibration.

[0118] The posture compensation instructions are input into the rotation execution component to control the posture calibration positioning. The rotation execution component is driven by an integrated servo motor and performs the rotation action of the branching wheel after receiving the compensation instructions. During the posture calibration positioning process, the system constantly maintains the elevation parameter of the contact surface of the branching wheel through a pneumatic regulating valve or a mechanical limiting device, so that the surface of the branching wheel is always higher than the plane of the roller by a certain distance. At the same time, high-precision angle sensors are used to collect dynamic deviation data of the rotation positioning, and the deviation data is transmitted back to the posture compensation mechanism optimization process in real time through a data bus. This transmission process forms a closed-loop control system, which dynamically adjusts the subsequent compensation calculation to improve the stability and adaptability of the system.

[0119] Specifically, the high-speed branching angle control method for the branching machine of the roller conveying line described in the present application includes the following steps:

[0120] The segment accumulation parameter and size data set of the branching machine are used as input data, and the length value of the toter is extracted from the input data as the dominant calculation weight factor;

[0121] The dominant calculation weight factor is input into the multi-parameter fusion operation process to generate an intermediate angle value;

[0122] The positioning offset correction parameter in the posture compensation mechanism is called, and the positioning offset correction parameter is combined with the intermediate angle value to generate a compensation control instruction;

[0123] Store the dominant weight factor distribution trajectory to the historical database, and update the multi-parameter fusion rule based on the distribution trajectory to adapt the left turning physical feature library.

[0124] The turning angle calculation optimization process disclosed by the application first takes the shunting machine section accumulation parameter and size data set as input data. The shunting machine section accumulation parameter includes device inherent structure parameters such as accumulation width and wheel spacing, and the size data set includes the length value and width value of the turnover box obtained by photoelectric scanning. The system standardizes the input format through the data preprocessing module, and establishes a unified physical quantity calculation benchmark.

[0125] The length value of the turnover box is extracted from the input data of the previous step as the dominant calculation weight factor. In the left turning angle calculation scenario, the length value is given a higher weight coefficient, and this design is based on the turning mechanics of long-size turnover boxes. The weight distribution module dynamically adjusts the influence of the factor according to the length-width ratio of the box, so that the calculation model is more consistent with the actual physical constraint conditions.

[0126] The dominant calculation weight factor is input into the multi-parameter fusion operation processing to generate a turning angle intermediate value. The fusion operation adopts a geometric constraint algorithm, and combines the accumulation width parameter to construct a turning force transmission model. The core algorithm calculates the best force direction of the shunting wheel through the vector decomposition principle, and outputs the uncorrected turning angle intermediate value. This intermediate value already includes the core influence of the dominant size parameter, but has not yet considered the deviation of the device operating state.

[0127] The positioning offset correction parameter in the posture compensation mechanism is called, and the positioning offset correction parameter is combined with the angle intermediate value of the previous step. The positioning offset correction parameter comes from real-time acquisition of dynamic positioning data of the shunting wheel, including mechanical gap deviation and thermal deformation compensation amount. The correction module adopts the vector superposition principle to perform angle vector compensation on the basis of the intermediate value to generate the final compensated control instruction. This instruction eliminates the influence of the cumulative error of the device on the turning accuracy.

[0128] The dominant weight factor distribution trajectory is stored to the historical database, and the multi-parameter fusion rule is updated based on the distribution trajectory. The system records the weight factor value and corresponding box size parameter in each calculation process, and identifies the parameter distribution law through the trajectory analysis module. When a specific size of the turnover box frequently appears, the fusion rule optimization instruction is triggered, and the constraint coefficient threshold of the left turning physical feature library is automatically adjusted. This adaptive mechanism continuously improves the scene adaptability of the calculation model.

[0129] Specifically, the adaptive shunting angle control method for the high-speed shunting machine of the drum conveying line disclosed by the application performs shunting wheel turning angle calculation based on length data, width data and accumulation parameters to generate a turning angle output value, and further includes:

[0130] The shunting machine section accumulation parameter and size data set are taken as calculation input data, the pallet width value is extracted from the input data of the previous step as a core calculation weight factor, and a rotation direction physical constraint model is constructed based on the core calculation weight factor;

[0131] The physical constraint model is input into multi-parameter fusion operation processing to generate an intermediate turning angle value;

[0132] The left turn physical compensation rule in the historical feature library is called, the left turn physical compensation rule is combined with the intermediate angle value of the previous step to generate a corrected control instruction;

[0133] Real-time turning trajectory feature data is captured through a trajectory monitoring device, and the trajectory feature data is compared with a preset tolerance model for verification processing, when the verification result is trajectory deviation, dynamic compensation parameters are generated according to the deviation, the dynamic compensation parameters are stored in a database and a rule iteration optimization instruction is triggered.

[0134] The turning angle calculation optimization process disclosed by the application first takes the shunting machine section accumulation parameter and size data set as calculation input data. The shunting machine section accumulation parameter includes structural parameters such as accumulation area width and wheel set spacing, and the size data set is composed of pallet length and width values provided by an optical-electrical scanning system. The system calls the two types of parameters through a data interface and performs standardization processing, establishes a unified calculation reference coordinate system, and provides physical environment constraint conditions for subsequent model construction.

[0135] The pallet width value is extracted from the input data of the previous step as a core calculation weight factor. This processing is aimed at the turning characteristics of pallets with large width dimensions, and gives the width parameter a higher calculation weight. The weight distribution engine dynamically adjusts the influence of the box length-width ratio factor, combined with the mechanical structure characteristics of the shunting machine, so that the width parameter plays a dominant role in the rotation direction physical constraint model.

[0136] A rotation direction physical constraint model is constructed based on the core calculation weight factor. The model construction module establishes a friction force distribution matrix according to the spatial layout parameters of the shunting wheel set and the contact area of the pallet bottom surface. By analyzing the influence mechanism of the width dimension on the turning resistance, a vector space equation describing the minimum turning resistance is generated, which limits the physical feasible range of the turning angle.

[0137] The physical constraint model is input into multi-parameter fusion operation processing to generate an intermediate turning angle value. The fusion operation uses a constraint optimization algorithm to solve the optimal turning angle in the physical feasible domain. The algorithm considers the pallet center of gravity offset trend and the shunting wheel driving force transmission efficiency, and outputs an intermediate angle value that meets the mechanical stability. The value has included the core influence of the width parameter, but has not yet fused the historical data of the equipment operation.

[0138] The left-turn physical compensation rule in the historical feature library is called, and the left-turn physical compensation rule is combined with the intermediate value of the angle in the previous step. The historical feature library stores successful case data of past left-turn operations, and the compensation rule extraction library extracts the steering deviation correction amount of similar size boxes. The compensation engine converts the correction amount into an angle vector and combines it with the intermediate value to generate a corrected control instruction. This process eliminates the influence of equipment mechanical clearance and transmission error on steering accuracy.

[0139] Real-time steering trajectory feature data is captured by a trajectory monitoring device. The monitoring device uses a laser displacement sensor array to collect motion trajectory coordinates in real time during the turning of the tote. The data processing module extracts feature values such as trajectory curvature rate of change and distance from the reference line to generate a set of trajectory feature data.

[0140] The trajectory feature data is compared with a preset tolerance model for verification processing. The tolerance model stores the standard turning trajectory parameter range of totes of different sizes, and the verification engine calculates the similarity index of the actual trajectory and the standard trajectory. When the similarity is lower than the set threshold, it is determined that there is a trajectory deviation event.

[0141] When the verification result is a trajectory deviation, a dynamic compensation parameter is generated according to the deviation. The parameter generation module analyzes the direction and amplitude of the trajectory deviation, and combines the current shunt wheel posture data to calculate the additional rotation correction amount. This dynamic compensation parameter is immediately applied to the subsequent shunt wheel control instruction to realize real-time correction of errors during operation.

[0142] The dynamic compensation parameter is stored in the database and triggers the rule iteration optimization instruction. The database uses a time series storage architecture to store compensation parameters, tote sizes, and environmental temperature and humidity data. When the cumulative number of similar size boxes reaches the statistical sample size, the system automatically triggers the rule optimization engine to update the left-turn physical compensation rule coefficients in the historical feature library, completing the adaptive evolution closed loop of the control model.

[0143] Specifically, the adaptive shunting angle control method for the high-speed shunting machine of the drum conveying line according to the present application comprises:

[0144] Real-time acquisition of the rotational speed state parameter of the shunt wheel driving unit as dynamic input data, and generation of a torque compensation instruction according to the fluctuation of the dynamic input data;

[0145] The torque compensation instruction is input into the friction contact surface pressure prediction model, and the friction contact surface pressure distribution trend is output through the pressure prediction model;

[0146] Based on the pressure distribution trend, the rotational output torque is dynamically adjusted, and the actual angle deviation of the shunt wheel is detected in real time;

[0147] When the actual angle offset exceeds the safety threshold, the following is performed: locking the current shunt wheel rotation trajectory, activating the emergency correction control process, generating state response instructions for the shunt wheel zero return action linkage, storing the dynamic parameter set of the conveying process and associating the positioning deviation data.

[0148] The process of conveying the turnover box by the shunt wheel surface friction described in the application first acquires the rotation speed state parameter of the shunt wheel driving unit as dynamic input data in real time. The acquisition process is realized by the Hall sensor installed on the output shaft of the servo motor, which monitors the real-time rotation speed fluctuation in the shunt wheel rotation process. The system generates a torque compensation instruction through a differential control algorithm according to the instantaneous change of the rotation speed data, which includes the torque adjustment amplitude and direction parameters, and is used to offset the driving force fluctuation caused by the load change.

[0149] The torque compensation instruction is input into the friction contact surface pressure prediction model, which is constructed based on the mapping relationship between the shunt wheel material characteristics and the friction coefficient of the turnover box bottom surface. After receiving the torque compensation instruction, the model combines the current shunt wheel steering angle parameter to calculate the change trend of the friction contact surface pressure distribution through the contact mechanics equation. The output result includes the pressure gradient change vector and the position of the maximum pressure point, which provides a theoretical basis for torque adjustment.

[0150] Based on the change trend of the pressure distribution, the rotation output torque is dynamically adjusted, and the control system adjusts the servo motor output through the vector frequency converter. The adjustment process follows the direction of the pressure gradient change to enhance the driving force transmission efficiency, and completes the torque redistribution within a microsecond response period. At the same time, a high-precision angle sensor is used to detect the actual angle offset of the shunt wheel in real time, and an angle offset trajectory data stream is generated, which is compared with the theoretical steering trajectory in real time.

[0151] When the actual angle offset exceeds the safety threshold, the safety response protocol is executed: first, trigger the trajectory locking module to lock the current shunt wheel rotation trajectory through the rigid brake; at the same time, activate the emergency correction control process, which includes the reverse rotation instruction generation module, output the state response instruction for the shunt wheel zero return action linkage; finally, store the dynamic parameter set of the conveying process, including the rotation speed fluctuation data, the pressure distribution matrix and the angle offset trajectory, and associate the positioning deviation data to construct the fault analysis data set. The whole process forms a closed-loop safety control mechanism, which guarantees the continuity of conveying while realizing the rapid disposal of abnormal state.

[0152] Specifically, the adaptive shunt angle control method for the high-speed shunt machine of the drum conveying line described in the application further comprises the following steps:

[0153] The motion posture type of the turnover box during the departure process is identified by the motion capture device, and the motion posture type is input into the reset timing classification model to generate a differentiated reset instruction.

[0154] synchronous scanning shunt section retention detection signal intensity parameter, when the retention signal intensity parameter exceeds the safety threshold, the auxiliary oscillation cleaning mechanism is started to process the retention, and an emergency channel switching control instruction is generated

[0155] correlate the process parameter of shunt wheel zero reset action with the historical database record, trigger the preventive maintenance evaluation process of the conveying equipment based on the correlation data, update the system event log and compress the dynamic parameter package to the maintenance database.

[0156] The processing process of the present application in response to the photoelectric trigger signal first identifies the motion posture type of the turnover box leaving process through the motion capture device. The device uses a high-speed industrial camera array combined with a machine vision algorithm to track the displacement trajectory of the turnover box edge profile, and identifies the vibration amplitude and deflection angle change in the box movement process. The motion posture type is divided into two characteristic modes of regular translation and irregular sliding, and the system extracts the feature vector as the input parameter of the reset timing classification model.

[0157] The motion posture type is input into the reset timing classification model to generate differentiated reset instructions. The classification model is constructed based on the support vector machine algorithm, and the training data includes various posture samples in the historical operation. The model output includes two types of instructions: when the posture type is regular translation, a fast reset instruction is generated to compress the shunt wheel zero reset time interval; when the posture type is irregular sliding, a buffer reset instruction is generated to prolong the reset action period to avoid mechanical impact. The instruction generation module synchronously outputs the servo motor acceleration curve parameters to realize flexible control of the shunt wheel reset action.

[0158] synchronous scanning shunt section retention detection signal intensity parameter, the process is realized by installing a laser scanning matrix in the shunt wheel gap. The matrix emits multiple beams of detection light, and the receiving end measures the reflection intensity decay rate and converts it into a signal intensity parameter. The system establishes a mapping relationship between the intensity parameter and the retention size, and sets a dynamic safety threshold range. When the signal intensity parameter continuously exceeds the upper limit of the threshold, it is determined as an effective retention event.

[0159] When the retention signal intensity parameter exceeds the safety threshold, the auxiliary oscillation cleaning mechanism is started to process the retention. The cleaning mechanism is driven by a piezoelectric ceramic vibration unit to generate a high-frequency micro-oscillation wave. The oscillation wave is transmitted to the retention contact surface through the shunt wheel support structure, which destroys the static friction balance state. At the same time, an emergency channel switching control instruction is generated, which includes a shunt wheel locking signal and a bypass conveying belt starting instruction, to establish a temporary shunt channel to ensure the continuity of the main line conveying.

[0160] The process parameters of the reset action of the associated shunt wheel are associated with the historical database record. The parameter association engine collects the servo motor current fluctuation curve, positioning accuracy deviation value and execution time parameter in the reset process, and performs feature matching with the historical operation record in the database. The matching process adopts a time series similarity algorithm to identify abnormal parameter combination patterns as device state evaluation inputs.

[0161] Based on the associated data, the preventive maintenance evaluation process of the conveying device is triggered. The evaluation process calls the device wear prediction model to analyze the shunt wheel bearing gap change trend and the driving gear meshing efficiency decay rate. When the key parameters approach the failure threshold, a hierarchical warning instruction is generated: the primary warning updates the system event log and prompts maintenance suggestions; the advanced warning triggers automatic shutdown protection, and at the same time, the dynamic parameter package is compressed and stored in the maintenance database. The database adopts an incremental storage strategy, retains the device running feature fingerprint throughout its life cycle, and provides a data basis for reliability analysis.

[0162] Specifically, the adaptive shunt angle control method for the high-speed shunt machine of the drum conveying line according to the present application, which synchronously executes the reset and zero of the shunt wheel and the new turnover box feeding process in response to the photoelectric trigger signal, further comprises:

[0163] Collect real-time safety state evaluation parameters of the shunt section as risk judgment inputs, and load the risk judgment inputs into the risk level judgment process to generate a risk level identifier;

[0164] According to the risk level identifier, perform a branch response operation, when the risk level identifier is low risk, generate a conveying unit quick start instruction, activate the pre-stage blocker release sequence, and reconstruct the conveying channel state matrix;

[0165] When the risk level identifier is high risk, start the safety buffer delay mechanism, and generate a set of channel defect repair instructions;

[0166] Associate the throughput counter value with the historical maintenance database record, iteratively optimize the risk level judgment threshold parameter based on the associated value, compress and store the state reconstruction process data package and update the maintenance log, and feed back the optimized threshold parameter to the risk level judgment process.

[0167] The safety control mechanism of the present application synchronously executes the reset of the shunt wheel and the new turnover box feeding process in response to the photoelectric trigger signal, first collects real-time safety state evaluation parameters of the shunt section as risk judgment inputs. The collection process is realized by a distributed sensor array, including a laser ranging unit, an infrared thermal imager and a vibration accelerometer installed in the conveying channel, which synchronously monitors the spatial obstacle distribution, the device temperature field change and the mechanical vibration frequency spectrum characteristics. The system fuses multi-source perception data into a safety state evaluation vector to provide real-time working condition input for risk level judgment.

[0168] The risk judgment input is loaded into a risk level judgment process to generate a risk level identifier. The judgment process uses a fuzzy logic decision engine to quantify the risk level of each parameter in the evaluation vector through a preset membership function. The engine outputs a binary risk level identifier: a low risk identifier indicates that the equipment is in a normal operating state; a high risk identifier indicates that there is a risk of mechanical interference or motion conflict. The identifier serves as the basis for decision-making for branch response operations, realizing digital expression of the risk situation.

[0169] The branch response operation is performed according to the risk level identifier: when the risk level identifier is low risk, a fast start instruction for the conveying unit is generated. The instruction triggers the control bus to send an enable signal to the conveying line master unit, synchronously activating the pre-segment blocker release sequence. The release sequence uses a timestamp triggering mechanism to gradually release the blocker lock state at preset time intervals, and reconstructs the conveying channel state matrix. The matrix update module reallocates the conveying path resource parameters, realizing the physical channel unobstructed for continuous box feeding.

[0170] When the risk level identifier is high risk, a safety buffer delay mechanism is started. The mechanism includes a time delay module and a space isolation module: the time delay module inserts a buffer time window to pause the new tote feeding process; the space isolation module generates a channel defect repair instruction set, which includes branch wheel emergency zero, conveying belt speed reduction, and sound-light alarm linkage operations, and cuts off the risk propagation path through physical isolation.

[0171] The throughput counter value is associated with the historical maintenance database record. The association engine uses a timestamp matching algorithm to establish a mapping relationship between the current throughput rate value and the device maintenance log in the historical database. The mapping relationship includes the device wear characteristic mode corresponding to a specific throughput rate interval, providing a statistical sample basis for threshold optimization.

[0172] Iterative optimization of risk level judgment threshold parameters based on associated values. The optimization engine uses an incremental learning strategy to analyze the correlation characteristics of fault events and throughput rate changes in historical maintenance records. When a significant increase in device abnormality rate at a specific throughput rate is detected, the risk judgment threshold parameter for the corresponding working condition is automatically lowered. The optimized threshold parameter is updated in real time to the fuzzy logic decision engine, completing the dynamic improvement of risk identification accuracy.

[0173] Compress the state reconstruction process data packet and update the maintenance log. The data compression module uses lossless coding technology to package the channel state matrix change record, blocker sequence operation timestamp, and risk event characteristic value into a binary data packet. The data packet is associated with the device serial number and time identifier and stored in the maintenance database, and the abnormal event classification label in the system maintenance log is updated, providing data support for device reliability analysis.

[0174] Specifically, the high-speed branching machine adaptive branching angle control method for the drum conveying line of the application includes:

[0175] Real-time acquisition of three-dimensional position offset dynamic parameters of the branching wheel as spacing control input, loading the spacing control input to the pressure distribution feedback model to generate pressure gradient parameters;

[0176] According to the safety state judgment of the pressure gradient parameters, when the pressure gradient parameters are within the safety threshold range, the high-frequency micro-vibration suppression balance period is started to stabilize the contact surface, and the rolling friction compensation calibration process is activated;

[0177] When the pressure gradient parameters exceed the safety threshold range, the rigid locking protection action of the trigger mechanism is triggered, and the mechanical deformation correction vector parameter is calculated;

[0178] Correlate the equipment running height parameter with the historical positioning database record to generate a calibration reference, dynamically reconstruct the pressure distribution feedback model parameters according to the calibration reference, generate the position offset accumulation trajectory parameter package and compress it to the maintenance database, and feed the stored parameters back to the pressure distribution feedback model optimization process.

[0179] The branching wheel friction conveying control process of the application first acquires three-dimensional position offset dynamic parameters of the branching wheel as spacing control input in real time. The acquisition process is realized by a laser displacement sensor array installed at the branching wheel support structure, which synchronously detects the micron-level position change of the branching wheel in X / Y / Z axial direction. The dynamic parameters include instantaneous displacement vector and acceleration component, which are converted into standard physical quantity format by the system to provide spatial reference input for pressure distribution modeling.

[0180] Load the spacing control input to the pressure distribution feedback model to generate the pressure gradient parameters. The model is constructed based on the Hertz contact theory, combined with the elastic modulus of the branching wheel and the material characteristics of the bottom surface of the turnover box, to establish the contact stress distribution equation. After receiving the three-dimensional position offset, the model outputs the pressure gradient parameters through the stress field simulation algorithm, including the maximum pressure point position and the stress change rate curve characteristics, representing the mechanical state change trend of the friction contact surface.

[0181] According to the safety state judgment of the pressure gradient parameters. The judgment engine compares the pressure gradient parameters with the preset dynamic safety threshold interval: when the parameters are within the safety threshold range, the high-frequency micro-vibration suppression balance period is started. The suppression process generates mechanical waves above 200Hz through piezoelectric ceramic actuators to offset the resonance effect of the contact surface; simultaneously activate the rolling friction compensation calibration process, which adjusts the branching wheel speed pulse width modulation parameters according to the pressure gradient direction to optimize the driving force transmission efficiency.

[0182] When the pressure gradient parameter exceeds the safety threshold, the rigid locking protection action of the mechanism is triggered. The protection action includes instantaneously energizing the electromagnetic brake to lock the shaft, and simultaneously calculating the mechanical deformation correction vector parameters. The calculation module analyzes the mapping relationship between the position offset data and the material yield strength, generating three-dimensional vector parameters including the correction torque and direction, driving the hydraulic compensation mechanism to perform plastic deformation repair.

[0183] The calibration benchmark is generated by associating the equipment's operating height parameters with historical positioning database records. The association engine calls upon height parameter records under the same operating conditions from the historical database and constructs a height-pressure characteristic curve using a time-series matching algorithm. The calibration benchmark dynamically updates the boundary constraints in the pressure distribution feedback model to eliminate systematic errors caused by equipment foundation settlement.

[0184] The pressure distribution feedback model parameters are dynamically reconstructed based on the calibration benchmark. The reconstruction process includes model coefficient updates and a verification loop: the coefficient update module replaces the stiffness matrix in the elastic contact equation; the verification loop verifies the model accuracy using real-time pressure sensor data and completes the adaptive iteration of the model parameters.

[0185] An accumulated trajectory parameter package of position offsets is generated and compressed for storage in the maintenance database. The parameter package uses a binary encapsulation format and includes a sequence of 3D position coordinates within a continuous time window, pressure gradient feature values, and a model reconstruction version identifier. The compression algorithm employs LZW lossless encoding technology to reduce storage resource consumption while preserving the integrity of the data features.

[0186] The stored parameters are fed back to the pressure distribution feedback model optimization process. The feedback loop establishes a real-time data channel between the database and the model engine. When a new type of turnover box material appears, the system automatically retrieves historical parameter packages for transfer learning and updates the friction coefficient mapping table in the material property parameter library, thereby achieving continuous evolution of pressure prediction capabilities.

[0187] Based on the appendix Figure 1 The control flow is as follows: After system initialization, the distribution wheel first swings to the straight-line position, simultaneously allowing the current station to accumulate boxes. The previous station's box exit action triggers the turnover box to enter the recognition area. Photoelectric sensors aggregate length and width data to form a size set. The host computer's conveying direction command is parsed into left distribution, right distribution, or straight-line type. The calculation unit performs steering angle calculations based on the distribution section's accumulation parameters. The left distribution uses a length-dominant algorithm, the right distribution prioritizes width constraint optimization, and the straight-line mode skips calculations and directly resets to zero. When the turnover box's front end reaches the distribution section, the position prediction algorithm generates trajectory data, the attitude compensation mechanism optimizes rotation accuracy, and the servo motor adjusts the distribution wheel angle to the target value in real time, maintaining a constant contact surface elevation to reduce friction interference. After the box is in place, the system verifies the next station's accumulation status based on the conveying direction. If box entry is allowed, the box exit and previous station's box entry actions are simultaneously initiated, compressing the cycle time and achieving continuous conveying.

[0188] By integrating servo drive and dynamic control logic, traditional lifting steps are eliminated, and the shunt wheel directly rotates to the calculated angle to complete steering or straight running, with compact and coherent action rhythm. During the friction conveying process, torque compensation instructions are dynamically adjusted based on speed fluctuations to output torque, and a pressure prediction model optimizes the contact surface pressure distribution to achieve stable steering of the turnover box; in response to the photoelectric trigger signal, the shunt wheel is synchronized to reset to zero and the new turnover box entry process is started, the motion posture classification model differentially controls the reset timing, and the risk of retention is scanned. The historical database iteratively updates the calculation rules and risk thresholds associated with equipment parameters to maintain system adaptability. The final solution improves conveying efficiency, supports high-speed line operation, avoids box jamming, and simplifies maintenance processes.

[0189] Please refer to Figure 2 , the flow direction block diagram of the high-speed shunt machine shows the multi-directional flow logic of goods in the drum conveying line, which includes three types of path topologies: straight running, left shunt, and right shunt. In straight running mode, goods enter the accumulation area of the high-speed shunt machine section from the straight line section of the previous station, are directly conveyed to the accumulation area of the next straight line section after a short buffer, and form a continuous straight channel. In the left shunt path, goods are transferred from the accumulation area of the high-speed shunt machine section to the left shunt triangular accumulation area, and are conveyed to the left shunt accumulation area of the next station after a steering buffer; the right shunt path follows a symmetric logic, and goods are conveyed to the right downstream station after steering through the right shunt triangular accumulation area. The triangular accumulation area, as a key steering and buffering unit, optimizes the centrifugal force distribution during steering of goods through geometric layout, maintaining conveying stability.

[0190] Path switching is achieved by real-time angle adjustment of the shunt wheel, without the need for traditional lifting mechanisms. When the conveying direction instruction is straight running, the shunt wheel maintains a 0° reference position, and goods seamlessly transition along the straight channel; the left shunt instruction triggers the shunt wheel to rotate to the calculated angle to the left, guiding the goods into the left triangular accumulation area; and the right shunt instruction drives the shunt wheel to symmetrically rotate to the right. The accumulation area state coordination mechanism achieves the continuity of path switching: before goods enter the accumulation area of the high-speed shunt machine section, the system verifies the allowed entry state of the next target accumulation area (straight running / left triangular / right triangular); only when the target accumulation area is ready, the goods are released at the current station and the entry permission of the previous station is triggered, eliminating the rhythm waiting gap. This design realizes the pipeline execution of the "entry-steering-exit" action of the conveying line.

[0191] The flow-to-frame is converted to physical conflict by triangular accumulation area. The traditional fixed angle branch is easy to cause cargo congestion due to lack of buffer. In this scheme, the triangular accumulation area as a dynamic expansion channel absorbs the inertial deviation in the conversion process. After the left branch triangular area receives the tangential force applied by the branch wheel, it gradually corrects the attitude of the goods through the trapezoidal layout; the right branch triangular area operates with mirror logic. The adaptive mechanism of branch wheel angle further optimizes the path efficiency: dynamically calculate the turning angle according to the size of the goods, avoid over-turning of small-size goods or insufficient turning of large-size goods, and realize the transportation of goods of different specifications along the path with the smallest resistance. Finally, a multi-directional conveying network with no interruption and high compatibility is formed, which supports continuous operation at a line speed of 1.4m / s.

[0192] Please refer to the attached Figure 3 , starting from the system initialization phase. The control logic first performs initialization operations, sets the branch wheel position to the straight direction reference state, and activates the box entry permission signal of the accumulation area at this station. The box exit action of the last accumulation area triggers the turnover box to enter the area of this station. The system synchronously retrieves the host computer instruction to obtain the conveying direction identifier (such as straight, left branch or right branch) and size information data of the current turnover box. This step realizes complete data collection, provides a basis for subsequent decision-making, and avoids delays or errors caused by data loss in traditional methods. The system verifies the state of the turnover box entering the identification area by continuously monitoring the photoelectric sensor signal, thereby seamlessly linking the conveying process and optimizing the action response time.

[0193] Based on the obtained size information and conveying direction identifier of the turnover box, the system performs the branch wheel swing angle calculation process. The calculation unit combines the branch machine section accumulation parameters such as width and wheel spacing, and applies the inverse trigonometric function model to generate the turning angle output value; the length-dominant algorithm is used for the left branch direction, the width optimization strategy is used for the right branch direction, and the zero angle instruction is directly output for the straight mode. Then, the system judges whether the accumulation area at this station is in place, and confirms the stable position of the turnover box through the photoelectric trigger signal. If the box is in place, the system analyzes the conveying direction identifier again, and verifies the box entry permission state of the next target accumulation area (the next station for straight, the triangular section for left branch, or the triangular section for right branch) according to the direction type (such as 1 for straight, 3 for left branch, and 4 for right branch). This decision logic eliminates waiting time, realizes compact action sequence, and improves overall beat efficiency.

[0194] When the target accumulation area is ready, the system synchronously executes the out-box action and the new-in-box process starts. Specifically, the station's turnover box out-boxes and leaves the station's photoelectric detection range at the same time, and immediately activates the station's last station in-box instruction to form a continuous conveying chain. The shunt wheel is then reset to the straight direction position, and the control process ends. During the process, the friction conveying mechanism maintains the stable turning of the turnover box by dynamically adjusting the rotation torque to avoid jamming; when the photoelectric trigger signal responds, the system optimizes the reset timing by associating the historical database to compress the beat gap. The final method supports high-speed line operation, realizes seamless switching of shunting and straight action, and enhances the system reliability and throughput capacity.

[0195] Shunting angle algorithm:

[0196] The straight direction is 0° for the shunting machine.

[0197] The shunting machine segment accumulation width is A, the distance between the first row of wheels and the third row of wheels of the shunting machine is C, the length of the turnover box is L, and the width is W, in mm. The shunting angle is α.

[0198] According to common use scenarios:

[0199]

[0200] After calculating the angle α, if the shunting direction is left shunting, the shunting wheel rotates to the left by an angle of α, and if the shunting direction is right shunting, the shunting wheel rotates to the right by an angle of α.

[0201] The shunting wheel angle α of the straight out-box is 0.

[0202] The present application solves the technical problems of adaptive turning and high-speed operation of the shunting machine through the following technical solutions: first, a dynamic angle calculation model is used to realize adaptive turning of the shunting wheel. The length and width data of the turnover box are collected in real time by the photoelectric array, and a rotation direction physical constraint model is constructed combined with the shunting machine segment accumulation parameters. The model performs differentiated calculation according to the conveying direction instruction type (left shunting, right shunting or straight line): for left shunting instruction, a length-dominated calculation process is called to generate a left turning angle; for right shunting instruction, a width-optimized calculation process is called to generate a right turning angle; and for straight line instruction, zero angle is directly output. During the calculation process, the posture compensation mechanism and the physical compensation rules of the historical feature library are fused to optimize the turning angle output value in real time, and the influence of mechanical deviation on turning accuracy is eliminated.

[0203] Secondly, the high-speed response control timing and structure optimization scheme is designed. When the totes arrive at the trigger position of the branching section, the system generates trajectory data through the position prediction algorithm to drive the integrated servo motor to execute the rotational positioning of the branching wheel. The branching wheel is installed at a constant height higher than the drum plane, reducing friction interference and achieving direct turning in the conveying process. The synchronous friction contact surface pressure prediction model is adopted to dynamically adjust the rotational output torque, and the high-frequency micro-vibration suppression technology is combined to stabilize the contact surface. When the totes leave the branching area, the system resets the branching wheel and the next tote feeding process through the photoelectric trigger signal linkage, compressing the action tempo.

[0204] Finally, a closed-loop safety optimization mechanism is established to ensure continuous high-speed operation. The system monitors pressure gradient parameters and position offset in real time and responds to abnormal states in stages: quickly rebuild the conveying channel state matrix in low-risk situations; trigger rigid locking protection and start channel repair in high-risk situations. Correlate the historical database to iteratively update the risk judgment threshold and pressure distribution model parameters, verify the turning effect through trajectory monitoring, and store dynamic compensation parameters. Based on the high-level parameters of equipment operation, a calibration benchmark is generated to continuously optimize the branching wheel turning angle calculation rules, forming a self-evolving control closed loop.

Claims

1. An adaptive branching angle control method for a high-speed branching machine in a roller conveyor line, characterized in that, include: Obtain the length, width, and conveying direction data of the turnover box; Call the branch unit segment accumulation parameters; Based on length data, width data, and accumulation parameters, the steering angle of the branch wheel is calculated, and a steering angle output value is generated. When the front end of the turnover box reaches the trigger position of the branch section, the control system drives the branch wheel to rotate to the steering angle output value; The turnover box is conveyed to the sorting completion state by friction on the surface of the sorting wheel; Real-time detection of photoelectric trigger signals when turnover boxes leave the sorting machine area; The system responds to photoelectric trigger signals to synchronously execute the branch wheel reset to zero and the new turnover box loading process.

2. The adaptive branching angle control method for a high-speed branching machine in a roller conveyor line according to claim 1, characterized in that, The process of obtaining the length data, width data, and conveying direction instructions of the turnover box includes: Receive the conveying direction command issued by the host computer as the direction control input; The length characteristics of the turnover box are scanned by a first photoelectric array arranged longitudinally along the roller conveyor line. The width distribution of the turnover box is captured by a second photoelectric array in the vertical conveying direction; Perform box posture scanning processing based on length characteristics and width distribution; When the box posture scan confirms that the turnover box has entered the recognition area, the aggregated length feature and width distribution form a set of size data. The size data set and direction control input are integrated into a transmission data packet, which is then transmitted to the control execution unit.

3. The adaptive branching angle control method for a high-speed branching machine in a roller conveyor line according to claim 2, characterized in that, The calculation of the steering angle of the branch wheels based on length data, width data, and accumulation parameters, and the generation of the steering angle output value, include: The accumulation parameters of the distribution machine section are obtained as the basic input data, and the length and width values ​​of the turnover box are extracted from the size data set as calculation parameters. The system identifies the type of transport direction command as a left branch command, a right branch command, or a straight-line command, and performs direction-adaptive calculations based on the identification results. When the command is identified as a left turn instruction, the calculation parameters are input into the length-dominant calculation process to generate the left turn angle; When the command is identified as a right-branch instruction, the calculation parameters are input into the width optimization calculation process to generate the right turn angle; When the command is identified as a straight-ahead command, the calculation process is skipped and a zero-angle command is output directly. Output the steering angle control command in the corresponding direction to the rotation control system, and update the calculation rules and adapt the parameter database based on the calculated parameters.

4. The adaptive branching angle control method for a high-speed branching machine in a roller conveyor line according to claim 3, characterized in that, When the front end of the turnover box reaches the trigger position of the branch section, the control system drives the branch wheel to rotate to the steering angle output value, including: Real-time capture of conveyor line displacement change parameters as displacement monitoring data; input displacement monitoring data into position prediction algorithm to generate position prediction trajectory. The position prediction trajectory is input into the attitude compensation mechanism to optimize the angle positioning accuracy, and the real-time attitude compensation command of the branch wheel is generated based on the optimization result. The attitude compensation command is input into the rotation execution component to control the attitude calibration and positioning. During the attitude calibration and positioning process, the elevation parameters of the contact surface of the branch wheel are kept constant. The dynamic deviation data of rotation positioning is collected and sent back to the attitude compensation mechanism optimization process.

5. The adaptive branching angle control method for a high-speed branching machine in a roller conveyor line according to claim 4, characterized in that, The step of calculating the steering angle of the branch wheel based on length data, width data, and accumulation parameters to generate a steering angle output value also includes: The accumulator parameters and size dataset of the shunting machine section are combined as input data, and the length of the turnover box is extracted from the previous input data as the dominant calculation weight factor. The dominant calculation weighting factor is input into the multi-parameter fusion calculation to generate the intermediate value of the steering angle; The positioning offset correction parameter in the attitude compensation mechanism is called, and the positioning offset correction parameter is combined with the intermediate value of the previous step angle to generate the compensated control command. Store the distribution trajectory of the dominant weighting factor to the historical database, and update the multi-parameter fusion rule based on the distribution trajectory to adapt to the left-turn physical feature library.

6. The adaptive branching angle control method for a high-speed branching machine in a roller conveyor line according to claim 5, characterized in that, The step of calculating the steering angle of the branch wheel based on length data, width data, and accumulation parameters to generate a steering angle output value also includes: The accumulation parameters and size datasets of the shunting machine section are combined as the calculation input data. The width value of the turnover box is extracted from the previous input data as the core calculation weight factor. A physical constraint model of the rotation direction is constructed based on the core calculation weight factor. The physical constraint model is input into a multi-parameter fusion calculation to generate an intermediate value for the steering angle. Call the left-turn physical compensation rule in the historical feature library, combine the left-turn physical compensation rule with the previous step angle median value, and generate the corrected control command; Real-time steering trajectory feature data is captured by a trajectory monitoring device. The trajectory feature data is compared with a preset tolerance model for verification. When the verification result is trajectory deviation, dynamic compensation parameters are generated based on the deviation amount. The dynamic compensation parameters are stored in the database and a rule iteration optimization instruction is triggered.

7. The adaptive branching angle control method for a high-speed branching machine in a roller conveyor line according to claim 6, characterized in that, The method of conveying the turnover box by friction of the distribution wheel surface includes: The rotational speed parameters of the branch wheel drive unit are collected in real time as dynamic input data, and torque compensation commands are generated based on the fluctuation of the dynamic input data. The torque compensation command is input into the friction contact surface pressure prediction model, and the pressure prediction model outputs the trend of friction contact surface pressure distribution. Dynamically adjust the rotational output torque based on the pressure distribution change trend, and detect the actual angular offset of the branch wheel in real time; When the actual angular offset exceeds the safety threshold, the following actions are executed: lock the current rotation trajectory of the guide wheel, activate the emergency correction control process, generate a status response command for the guide wheel zeroing action linkage, store the dynamic parameter set of the conveying process and associate it with the positioning deviation data.

8. The adaptive branching angle control method for a high-speed branching machine in a roller conveyor line according to claim 7, characterized in that, The response photoelectric trigger signal also includes: The motion capture device identifies the motion posture type of the turnover box during the departure process, and the motion posture type is input into the reset timing classification model to generate differentiated reset commands. The system synchronously scans the signal strength parameters of obstructions in each road segment. When the signal strength parameter exceeds a safety threshold, the auxiliary oscillation clearing mechanism is activated to handle the obstruction, generating an emergency lane switching control command. The process parameters of the associated shunt wheel zeroing action are linked to historical database records. Based on the associated data, the preventive maintenance assessment process of the conveyor equipment is triggered, the system event log is updated, and the dynamic parameter package is compressed and stored in the maintenance database.

9. The adaptive branching angle control method for a high-speed branching machine in a roller conveyor line according to claim 8, characterized in that, The synchronous execution of the branch wheel reset and zeroing process and the start of the new turnover box entry process in response to the photoelectric trigger signal also includes: Collect real-time safety status assessment parameters for each road segment as input for risk assessment, and load the risk assessment input into the risk level assessment process to generate risk level identifiers; Branch response operations are executed based on the risk level identifier. When the risk level identifier is low risk, a fast start command for the conveyor unit is generated, the release sequence of the front-end blocker is activated, and the state matrix of the conveyor channel is reconstructed. When the risk level is identified as high risk, a safety buffer delay mechanism is activated to generate a set of channel defect repair instructions; The throughput counter values ​​are correlated with historical maintenance database records. The risk level determination threshold parameters are iteratively optimized based on the correlated values. The data packets of the storage state reconstruction process are compressed and the maintenance logs are updated. The optimized threshold parameters are then fed back to the risk level determination process.

10. The adaptive branching angle control method for a high-speed branching machine in a roller conveyor line according to claim 9, characterized in that, The step of conveying the turnover box to the sorting completion state through the friction of the sorting wheel surface includes: The dynamic parameters of the three-dimensional position offset of the branch wheel are collected in real time and used as the spacing control input. The spacing control input is then loaded into the pressure distribution feedback model to generate pressure gradient parameters. Based on the pressure gradient parameters, a safety status determination is performed. When the pressure gradient parameters are within the safety threshold range, the high-frequency micro-vibration suppression balance cycle is activated to stabilize the contact surface and the rolling friction compensation calibration process is initiated. When the pressure gradient parameter exceeds the safety threshold range, the rigid locking protection action of the mechanism is triggered, and the mechanical deformation correction vector parameter is calculated. The system generates a calibration benchmark by associating the equipment's operating altitude parameters with historical positioning database records. Based on the calibration benchmark, it dynamically reconstructs the parameters of the pressure distribution feedback model, generates a position offset accumulation trajectory parameter package, compresses and stores it in the maintenance database, and feeds the stored parameters back to the pressure distribution feedback model optimization process.

Images