Remote control system for an iron accessory end product detection sorting machine

Abstract

The application relates to the technical field of industrial automation control, and discloses a remote control system of an iron accessory finished product detection sorting machine, which comprises the iron accessory finished product detection sorting machine, a local control system and a remote control system. A data acquisition module acquires initial pulse values and system time stamps to construct a space-time joint identifier and uploads detection data. A pulse reconstruction module calculates a local pulse equivalent in combination with a trigger state of a position verification sensor, generates a global dynamic pulse equivalent through filtering, adaptively updates a target execution pulse, and realizes mechanical slip compensation. In the working condition of receiving a new detection threshold, a routing execution module cuts into a double-buffering state, locks in-transit gray area workpieces, executes secondary logic comparison in parallel by using the new and old threshold values, and allocates an auxiliary isolation channel for critical workpieces with inconsistent judgment results. The application solves the conveying displacement error generated by equipment operation, avoids the misclassification phenomenon in the critical period of parameter hot updating, and guarantees the stability of continuous system operation.

Description

Technical Field

[0001] This invention relates to the field of industrial automation control technology, specifically a remote control system for a finished iron accessory inspection and sorting machine. Background Technology

[0002] In the manufacturing process of power fittings and iron accessories, the inspection of finished product appearance quality and the sorting of defective products are essential steps in controlling the product qualification rate. With the popularization of the Industrial Internet, sorting equipment is gradually incorporating machine vision inspection and remote control technologies to achieve centralized management of quality inspection parameters. Conventional control schemes typically rely on the static pulse equivalent output by the spindle encoder to calculate the spatial position of the workpiece on the conveyor belt and control the sorting actions of the downstream actuators accordingly.

[0003] However, in actual continuous production environments, conveyor belts experience mechanical wear, material elastic deformation, and load variations due to long-term operation, leading to mechanical slippage between the bottom of the workpiece and the conveyor belt surface. This slippage causes errors in the coordinates calculated based on static pulse equivalents, easily resulting in offsets in the trigger position of sorting actions.

[0004] Furthermore, when sorting equipment is connected to a remote control architecture, there is an asynchronous network communication delay between the local system and the remote server. If, during continuous equipment operation, the remote control system dynamically issues a new quality inspection threshold, the workpieces en route on the conveyor belt, located between the inspection station and the sorting execution mechanism, will face a sudden change in the judgment benchmark. Since the initial state data of these workpieces has already been collected, directly applying the new threshold for judgment will cause sorting logic errors in workpieces at the boundary between old and new parameters. To avoid such erroneous actions caused by changes in process standards, existing technologies typically employ a method of stopping the machine, emptying the conveyor belt, and then writing the new parameters. This approach interrupts the continuous operation cycle of the production line, reduces the overall efficiency of the equipment, and fails to meet the coordinated requirements of modern production lines for continuous equipment operation and hot parameter updates. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a remote control system for a finished iron accessory inspection and sorting machine. This system solves the problems of shifted sorting action trigger positions caused by mechanical slippage during long-term conveyor belt operation, and sorting logic errors in transit caused by the alternation of old and new standards when remotely updating detection thresholds during continuous equipment operation.

[0006] To achieve the above objectives, the present invention provides the following technical solution: A remote control system for a finished iron accessory inspection and sorting machine includes: The finished iron accessory inspection and sorting machine is used to carry and transport the workpieces to be tested, provide trigger signals and perform sorting actions. The finished iron accessory inspection and sorting machine includes a conveyor belt, a spindle absolute encoder, an inspection station, a position verification sensor, a main rejection actuator and an auxiliary isolation actuator. A local control system is used to connect to the finished iron accessory inspection and sorting machine. The local control system includes a data acquisition module, a pulse reconstruction module, and a routing execution module. A remote control system for communicating with the local control system, the remote control system including a command issuing and parameter management module; The data acquisition module is used to generate a spatiotemporal joint identifier for the workpiece under test based on the output of the detection station and the spindle absolute encoder, and upload the detection data to the remote control system. The instruction issuance and parameter management module is used to generate sorting instructions based on the detection threshold and return them to the local control system. The pulse reconstruction module is used to calculate and update the target execution pulse of the workpiece under test by combining the trigger state of the position verification sensor; The routing execution module is used to allocate sorting routes and output electrical drive signals to the main rejection execution mechanism or the auxiliary isolation execution mechanism.

[0007] Preferably, the detection station is equipped with a trigger photoelectric switch and an image acquisition device; When the workpiece under test enters the detection station and blocks the trigger photoelectric switch, the data acquisition module extracts the system timestamp, latches the initial pulse value output by the spindle absolute encoder, and generates a serial number. The data acquisition module combines the sequence number, the system timestamp, and the initial pulse value to construct the spatiotemporal joint identifier. The data acquisition module encapsulates the detection data obtained by the image acquisition device and the spatiotemporal joint identifier into a network message and uploads it to the remote control system.

[0008] Preferably, the position verification sensor is located downstream of the detection station; When the workpiece under test triggers the position verification sensor, the pulse reconstruction module extracts the actual pulse value at the trigger moment; The pulse reconstruction module extracts the initial pulse value from the spatiotemporal joint identifier; The local control system reads the first distance data from the detection station to the position verification sensor; The pulse reconstruction module combines the actual pulse value, the initial pulse value, and the first distance data to calculate and obtain the local pulse equivalent.

[0009] Preferably, the pulse reconstruction module has an internal first-in-first-out data queue; The pulse reconstruction module determines the address usage status of the first-in-first-out data queue; If the first-in-first-out data queue is not full, the pulse reconstruction module appends the local pulse equivalent to the tail of the first-in-first-out data queue. The pulse reconstruction module extracts all existing historical local pulse equivalent data in the first-in-first-out data queue, performs filtering operations, and generates a global dynamic pulse equivalent.

[0010] Preferably, the local control system reads the second distance data from the detection station to the main rejection actuator; The pulse reconstruction module uses the global dynamic pulse equivalent and the second distance data, combined with the initial pulse value, to perform adaptive compensation calculation; The pulse reconstruction module calculates and updates the target execution pulse corresponding to the workpiece under test through the adaptive compensation calculation. The local control system writes the target execution pulse into the waiting execution queue.

[0011] Preferably, the instruction issuance and parameter management module receives the network packet, performs logical operations on the detection data according to the currently effective detection threshold, and generates the sorting instruction containing the judgment result and the spatiotemporal joint identifier; When the instruction issuance and parameter management module issues the sorting instruction containing the new detection threshold, the local control system allocates an independent address space to write the new detection threshold. The routing execution module switches to double-buffered mode.

[0012] Preferably, after the routing execution module switches to the double-buffered state, the local control system reads the initial pulse value and the target execution pulse corresponding to the in-transit workpiece in the waiting execution queue; The local control system continuously reads the current pulse value of the spindle absolute encoder and extracts the in-transit workpieces whose current pulse value is greater than or equal to the initial pulse value and less than the target execution pulse to construct a gray zone workpiece set. The routing execution module locks the gray zone workpieces within the gray zone workpiece set.

[0013] Preferably, the local control system is set with an initial global detection threshold and pre-stores the original feature scalars of the workpiece under test; The routing execution module extracts the original feature scalar corresponding to the gray area workpiece; The routing execution module performs a secondary logical comparison between the original feature scalar and the initial global detection threshold and the new detection threshold, respectively. The routing execution module independently outputs a first decision Boolean value based on the initial global detection threshold and a second decision Boolean value based on the new detection threshold for a single gray area workpiece.

[0014] Preferably, the routing execution module assigns the corresponding sorting route to the gray zone workpiece based on the first determination Boolean value and the second determination Boolean value; The sorting route includes a release channel, a main rejection channel, and an auxiliary isolation channel; If the first and second Boolean values ​​are inconsistent, the routing execution module allocates the auxiliary isolation channel to the gray zone workpiece.

[0015] Preferably, the local control system continuously reads the current pulse value of the spindle absolute encoder; When the current pulse value is greater than or equal to the target execution pulse, the routing execution module outputs the electrical drive signal to the main rejection execution mechanism or the auxiliary isolation execution mechanism to perform the corresponding action according to the assigned sorting route; After the corresponding action is completed, the local control system releases the memory space occupied by the gray area workpiece, and deletes the initial global detection threshold and exits the double buffer state after the number of gray area workpieces is reduced to zero.

[0016] This invention provides a remote control system for a finished iron accessory inspection and sorting machine. It has the following advantages: 1. This invention calculates and updates the target execution pulse of the workpiece by combining data from a pulse reconstruction module and a position verification sensor. The system introduces the local pulse equivalent into a first-in-first-out data queue for filtering and calculation to generate a global dynamic pulse equivalent. This adaptive compensation mechanism eliminates slippage errors caused by conveyor belt mechanical wear, elastic deformation, and load changes, ensuring the position triggering accuracy of the sorting mechanism under long-term continuous operation conditions.

[0017] 2. This invention employs a data acquisition module to generate a spatiotemporal joint identifier for the workpiece under test, and encapsulates the serial number, system timestamp, and initial pulse value at the hardware level. This structured identifier design, combined with asynchronous network message interaction, enables the local control system to continue executing the underlying cyclic scanning after uploading data to the remote control system, avoiding the blocking interference of network communication delays on the underlying high-speed mechanical conveying cycle, and achieving decoupling of displacement and data calculation.

[0018] 3. This invention utilizes a double-buffering mechanism in the routing execution module to handle the dynamic distribution of detection thresholds. The system can lock the workpiece in the gray zone between the detection station and the execution mechanism without stopping the system, and perform a secondary logical comparison of the original feature scalars using the initial global detection threshold and the new detection threshold. For critical workpieces where changes in old and new process standards result in inconsistent judgments, the system independently allocates an auxiliary isolation channel to prevent misjudgments and misclassifications during the hot update of production line parameters. Attached Figure Description

[0019] Figure 1 This is a system architecture diagram of the present invention; Figure 2 This is a schematic diagram of the method flow of the present invention; Figure 3 This is a flowchart of the calibration and initialization process of the present invention; Figure 4 This is a flowchart of the detection triggering and multi-dimensional state quantity latching mechanism of the present invention; Figure 5 This is a flowchart of the secondary position verification and local pulse parameter extraction process of the present invention; Figure 6 This is a flowchart of the queue management and global dynamic pulse reconstruction process of the present invention; Figure 7 The flowchart shows the adaptive correction process of the target execution pulse in this invention. Figure 8 This is a flowchart of the asynchronous instruction reception and system state machine switching process of the present invention; Figure 9 The graph shows a comparison of the effects of the dynamic pulse reconstruction adaptive compensation mechanism of the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] See attached document Figure 1This invention provides a remote control system for a finished iron accessory inspection and sorting machine, comprising: Finished iron accessories inspection and sorting machine, local control system and remote control system.

[0022] The finished iron accessory inspection and sorting machine is used to carry and transport the workpieces to be inspected, and provides trigger signals at preset cycle points to execute corresponding sorting actions. The finished iron accessory inspection and sorting machine includes a conveyor belt, a spindle absolute encoder, an inspection station, a position calibration sensor, a main rejection actuator, and an auxiliary isolation actuator.

[0023] The spindle absolute encoder is coaxially connected to the drive spindle of the conveyor belt and is used to output the absolute pulse value of the conveyor belt during its movement.

[0024] The inspection station is located at the beginning of the conveyor belt, and is equipped with a trigger photoelectric switch and an image acquisition device.

[0025] The position verification sensor is set at a fixed distance downstream of the detection station to obtain the offset reference point of the workpiece as it actually moves on the conveyor belt.

[0026] The main rejection actuator is located downstream of the position verification sensor and is used to perform the rejection action of defective workpieces.

[0027] The auxiliary isolation actuator is located adjacent to the main rejection actuator and is used to perform isolation and diversion actions on workpieces that are in a critical state of parameter change.

[0028] The local control system is deployed on the production floor and is connected to all electrical components of the finished iron accessory inspection and sorting machine via hardwire. The local control system includes a data acquisition module, a pulse reconstruction module, and a routing execution module.

[0029] The remote control system is deployed on a remote server and includes a command issuance and parameter management module. The local control system and the remote control system communicate asynchronously via industrial Ethernet.

[0030] See attached document Figure 2 This invention provides a remote control method for a finished iron accessory inspection and sorting machine. Based on the aforementioned hardware architecture, the method includes the following steps: The system is powered on and executes the initialization calibration program. The local control system reads the first distance data from the detection station to the position verification sensor and the second distance data from the detection station to the main rejection actuator. The local control system sets the initial standard pulse equivalent and the initial global detection threshold.

[0031] The iron accessory workpiece enters the inspection station along the conveyor belt and blocks the photoelectric switch. The data acquisition module of the local control system extracts the current system timestamp at the moment of trigger edge and latches the initial pulse value output by the spindle absolute encoder.

[0032] The data acquisition module generates a serial number for the workpiece. It then combines the serial number, system timestamp, and initial pulse value to construct a spatiotemporal joint identifier, which the image acquisition device uses to obtain the detection data. Finally, the data acquisition module encapsulates the detection data and the spatiotemporal joint identifier into a network message, which is then uploaded to the remote control system via industrial Ethernet.

[0033] The workpiece moves along the conveyor belt and triggers the position verification sensor. The pulse reconstruction module of the local control system extracts the real pulse value at the trigger moment. The pulse reconstruction module extracts the initial pulse value in the spatiotemporal joint identifier and performs local pulse equivalent calculation based on the pulse difference and the first distance data.

[0034] The pulse reconstruction module uses the built-in first-in-first-out data queue to perform filtering operations on the local pulse equivalent to obtain the global dynamic pulse equivalent. The pulse reconstruction module uses the global dynamic pulse equivalent and the second distance data to calculate and update the target execution pulse of the workpiece.

[0035] The remote control system's instruction issuance and parameter management module receives network messages, performs logical operations on the detection data based on the currently effective detection threshold, and generates a sorting instruction containing the judgment result and the original spatiotemporal joint identifier, which is then returned to the local control system via industrial Ethernet.

[0036] The local control system receives sorting instructions. When the instruction is issued and the parameter management module issues a new detection threshold, the routing execution module enters a double-buffered state. The routing execution module locks the workpiece located in the gray zone between the detection station and the main rejection mechanism.

[0037] For gray zone workpieces, the routing execution module performs a secondary logical comparison by calling the initial global detection threshold and the new detection threshold in parallel. Based on the logical comparison result, the routing execution module assigns a corresponding sorting route to the gray zone workpiece. The sorting route includes a release channel, a main rejection channel, and an auxiliary isolation channel.

[0038] The local control system's underlying scanning program continuously reads the current pulse value of the spindle absolute encoder and compares the current pulse value with the target execution pulse.

[0039] Under the condition that the current pulse value is greater than or equal to the target execution pulse, the routing execution module outputs an electrical drive signal to the main rejection execution mechanism or the auxiliary isolation execution mechanism to perform the corresponding action according to the assigned sorting route. After the action is completed, the local control system releases the memory space occupied by the workpiece. After the number of workpieces in the gray area is reduced to zero, the local control system deletes the initial global detection threshold and exits the double buffer state.

[0040] See attached document Figure 3 Before acquiring the state of the workpiece under test and generating a spatiotemporal joint identifier, the local control system needs to complete the digital mapping of the equipment operating parameters.

[0041] The calibration and initialization process is performed after the sorting machine is powered on or after a fault reset, and specifically includes the following steps: S201, construct a one-dimensional spatial coordinate system and load the device position distance parameters.

[0042] The local control system takes the vertical plane where the optical axis of the photoelectric switch triggered inside the detection station is located as the origin of the one-dimensional coordinate system. The local control system acquires the first distance data from the detection station to the position verification sensor, and the second distance data from the detection station to the center action point of the main rejection actuator.

[0043] The methods for obtaining the aforementioned first and second distance data include the following specific implementation approaches: On-site operators manually input parameters through the parameter setting window of the local human-machine interface and send them to the controller; alternatively, the background resident program of the local control system reads the pre-set equipment engineering configuration file in the memory for parameter parsing. The aforementioned equipment engineering configuration file is specifically an Extensible Markup Language (XML) format file or a JavaScript object-based JSON format file.

[0044] S202, calculate and load the initial standard pulse equivalent.

[0045] The initial standard pulse equivalent represents the number of pulses output by the spindle absolute encoder for every millimeter the conveyor belt moves in the running direction. The local control system derives this specific value based on the rated total number of pulses per revolution of the spindle absolute encoder, combined with the outer diameter of the main drive roller and the transmission ratio of the mechanical reducer between the connecting motor and the main drive roller, through circumference division and transmission ratio conversion.

[0046] For the selection of the spindle absolute encoder model and the hardware data reading interface, those skilled in the art can read its values ​​using the conventional SSI protocol or PROFINET industrial Ethernet bus according to the on-site electrical environment. The wiring method and communication configuration are well-known technologies in the field and will not be described in detail here.

[0047] S203, configures the queue memory space for pulse equivalent dynamic reconstruction operations.

[0048] The local control system allocates a contiguous address space in its internal random access memory to construct a first-in-first-out (FIFO) data queue. The local control system reads a preset queue depth parameter. Based on this queue depth parameter, the local control system determines the maximum length limit of historical local pulse equivalent values ​​that the FIFO data queue can store.

[0049] S204, Load the initial global detection threshold for the current batch of products.

[0050] The local control system retrieves the quality inspection formula data from the previous shift from the local non-volatile memory, or receives the initial parameter synchronization network message issued by the remote control system at the initial stage of connection establishment. The local control system parses the above data stream and extracts the initial global detection threshold.

[0051] To support the actual quality inspection logic, the initial global inspection threshold includes specific physical quantity judgment standards for different inspection dimensions. Specifically, the initial global inspection threshold includes the upper limit of the allowable area of ​​scratches on the surface of the iron accessory and the upper and lower deviation tolerance values ​​of the outer contour dimensions of the iron accessory. The local control system extracts and assigns the values ​​to the global variable register designated by the system. The value in this register will serve as the sole legal judgment benchmark in the current continuous operation state of the production line until a new parameter change instruction is received.

[0052] See attached document Figure 4 After parameter initialization is completed, the system enters continuous production operation mode and performs parallel capture of state variables when the workpiece reaches the detection origin.

[0053] The detection triggering and multidimensional state quantity latching mechanism specifically includes the following steps: S205, captures the trigger event of the detection station.

[0054] The iron accessory workpiece is placed on the conveyor belt and moves towards the inspection station. When the workpiece passes the origin and blocks the detection beam of the photoelectric switch, the output of the photoelectric switch generates a level transition signal. The high-speed digital input channel of the local control system receives this level transition signal and configures it as a hardware interrupt source to trigger the underlying logic.

[0055] For the optical selection, installation position adjustment and electrical wiring method of the trigger photoelectric switch, those skilled in the art can make selections according to the actual working conditions on site. The signal jitter removal and hardware filtering processing are well known technologies in this field and will not be described in detail here.

[0056] S206 responds to a hardware interrupt and retrieves the system timestamp.

[0057] Upon detecting the rising or falling edge of the aforementioned hardware interrupt, the central processing unit (CPU) of the local control system immediately suspends the main program task of the current loop scan and jumps to the interrupt handling routine. The CPU calls the low-level interface of the internal real-time clock module to read the current clock register value and generate the system timestamp of the workpiece entering the inspection station.

[0058] The specific characteristics of the above system timestamp are absolute timestamp format data accurate to the millisecond level, or the microsecond-level periodic count value accumulated by the local control system since power-on.

[0059] S207, synchronously latches the initial pulse value of the spindle absolute encoder.

[0060] The conveyor belt operates continuously at high speed during normal production. To eliminate position lag errors caused by software polling, the local control system performs hardware-level data latching in the interrupt handling routine. Upon receiving a signal indicating a level transition from the same source, the high-speed counting module built into the local control system freezes the currently accumulated spindle absolute encoder value in its internal pulse register. The local control system reads this frozen register value and uses it as the initial pulse value for the current workpiece.

[0061] To ensure strict alignment between the information domain and the physical domain data, the local control system binds the system timestamp generated by the central processing unit clock to the initial pulse value of the spindle absolute encoder, which is hardware-latched by the high-speed counting module, for each workpiece passing sequentially on the conveyor belt. The system encapsulates this time scalar and displacement scalar into a unified initial state set to ensure spatiotemporal consistency for subsequent logic tracking.

[0062] S208 sends out image acquisition trigger commands in parallel.

[0063] Within the same control cycle after the timestamp and pulse value latch are completed, the local control system sends a hardwired trigger pulse to the image acquisition device inside the detection station through a specific digital output channel. After receiving the trigger pulse, the image acquisition device performs camera exposure and data acquisition actions.

[0064] The above actions ensure that the visual feature data acquired by the image acquisition device and the recorded initial pulse value are at the same zero-point reference plane in spatial location, thus providing a closely corresponding data source for subsequent spatiotemporal identifier assembly.

[0065] After acquiring the initial state quantities of the workpiece under test, the local control system needs to structurally integrate the discrete hardware scalars and establish a non-blocking communication link with the remote control system. The structured encapsulation and asynchronous interaction process of this spatiotemporal joint identifier specifically includes the following steps.

[0066] S209 assigns a unique serial number to the workpiece that is currently triggering the detection.

[0067] The local control system maintains a global counter variable. Each time a hardware interrupt is responded to and a trigger event is captured, the local control system performs an increment operation on the global counter variable, generating an incrementing unsigned long integer value, and uses this value as the sequence number of the current workpiece.

[0068] For the generation of identifiers in distributed architecture systems, those skilled in the art can use globally unique identifier algorithms. The encoding rules and hash calculation process are well-known technologies in the field and will not be elaborated here.

[0069] S210, Construct a spatiotemporal joint identifier data structure.

[0070] The local control system's memory management module allocates space for a data structure of a specific length. The local control system combines the generated sequence number (used to provide a unique index key value in network packets), the system timestamp representing the time origin latched in the previous logic process, and the initial pulse value characterizing the workpiece's spatial starting coordinates, into a spatiotemporal joint identifier. This data structure achieves the cascading fusion of spatial displacement, system clock reference, and information logic identifier within the same memory block.

[0071] S211, assemble and send asynchronous network interaction messages.

[0072] The local control system extracts the original inspection data of the current workpiece from the data bus of the inspection station. The aforementioned original inspection data specifically includes scalar data of defect features output by the vision algorithm of the intelligent camera, or array of geometric dimension features output by the measurement grating. The local control system uses the original inspection data as the load data area and the spatiotemporal joint identifier as the message header identification area. The two are spliced ​​and assembled into an application layer network message.

[0073] The local control system invokes the Transmission Control Protocol (TCP) socket interface of the Ethernet protocol stack to send the network packet to the remote control system. After the network packet transmission command is issued, the communication thread of the local control system immediately returns and continues executing the main program's loop scanning task, without blocking and waiting for the remote control system's confirmation of receipt and return of calculation results. This non-blocking low-level communication call constitutes the asynchronous data interaction mechanism of this system.

[0074] S212, Establish a local waiting queue for execution.

[0075] While executing network transmission actions, the local control system allocates a separate data list area and a local feature cache area in its internal memory. The local control system pushes a copy of the aforementioned spatiotemporal joint identifier into this data list area, forming a queue waiting to be executed.

[0076] Simultaneously, the local control system stores the original detection data or feature scalars of the current workpiece in the local feature cache and establishes a binding mapping with the sequence number in the spatiotemporal joint identifier. The system marks the status bit of this record in the queue as suspended. This queue is used to continuously track the coordinate reference of the workpiece locally and wait for subsequent instructions from the remote control system to unsuspend and update the status.

[0077] See attached document Figure 5 In industrial environments where workpieces are continuously transported, unpredictable mechanical slippage can occur between the conveyor belt and the bottom of the workpiece due to load variations or mechanical wear.

[0078] To eliminate the coordinate offset caused by the mechanical slip, the local control system performs a secondary position check and extracts local pulse parameters before executing the action.

[0079] The process specifically includes the following steps: S301, captures the trigger event of the position verification sensor and latches the verification pulse value.

[0080] After the workpiece leaves the inspection station with the conveyor belt, it continues to move. When its physical object blocks the position verification sensor, the position verification sensor outputs a level change signal, which is captured by the high-speed input channel of the local control system.

[0081] The high-speed counting module inside the local control system responds to the signal, latches the real pulse value output by the spindle absolute encoder at this moment, and records it as the verification pulse value. The specific lower-level feature of the above position verification sensor is a high-precision laser beam sensor or a diffuse reflection photoelectric sensor.

[0082] For signal filtering and jitter reduction processing of position verification sensors, those skilled in the art can use conventional RC hardware filtering or set software dead time. The signal jitter reduction processing is a well-known technology in the field and will not be described in detail here.

[0083] S302, traverse the waiting execution queue and extract the initial state data.

[0084] After reading the aforementioned level transition signal, the local control system accesses the waiting execution queue established in the internal memory. According to the order in which the workpieces enter the detection station, the local control system searches for the data record that is in the suspended state and is at the top of the queue. The local control system extracts the spatiotemporal joint identifier corresponding to the record. The system parses and reads the initial pulse value recorded at the detection origin of the workpiece from the data structure of the extracted spatiotemporal joint identifier.

[0085] S303 performs kinematic operations to obtain the local pulse equivalent.

[0086] The processor of the local control system calls the first distance data loaded during the system initialization phase. Combining this with the extracted initial pulse value and verification pulse value, the local control system calculates the real-time pulse feedback parameters for the current workpiece segment through division. The formula for this calculation is as follows: ; in, This represents the calculated local pulse equivalent, which indicates the actual pulse count value per millimeter within this section of the conveyor belt under the current load condition. This indicates the value of the verification pulse that is instantly latched by the workpiece trigger position verification sensor. This indicates the initial pulse value generated when the workpiece is triggered at the inspection station; This represents the first distance data calibrated by the system. Through the above calculations, the system converts the spatial sliding displacement error of the individual workpiece on the conveyor belt into specific digital equivalent parameters.

[0087] See attached document Figure 6 After acquiring the local pulse parameters of a single workpiece, the system needs to eliminate the interference of environmental noise and perform global smoothing on the pulse parameters.

[0088] The process of queue management and global dynamic pulse reconstruction specifically includes the following steps: S304 checks the memory status and performs a local pulse equivalent push-up update operation.

[0089] The local control system reads the local pulse equivalent obtained in the previous calculation process, accesses the pre-configured first-in-first-out data queue in the internal random access memory, and reads the current address usage status of the queue.

[0090] The local control system determines whether the number of data items stored in the current queue has reached the queue depth parameter set during system initialization. If the queue is not full, the local control system directly appends the local pulse equivalent to the tail of the queue. If the queue has reached the maximum storage depth, the local control system executes a memory overwrite instruction, discarding the oldest historical pulse equivalent data written at the head of the queue and pushing the latest local pulse equivalent to the bottom of the queue.

[0091] For the underlying memory addressing logic and read / write pointer state switching of the first-in-first-out (FIFO) data queue, those skilled in the art can use a conventional circular buffer array structure for programming implementation. The specific addressing process is a well-known technology in this field and will not be described in detail here.

[0092] S305 executes a filtering algorithm to reconstruct the global dynamic pulse equivalent.

[0093] Within the same control cycle after completing a single data push operation, the local control system initiates a data traversal task. In order to eliminate outlier numerical interference caused by the manufacturing tolerance of a single workpiece surface or the jitter of a single electrical signal from a position verification sensor, the local control system extracts all existing historical local pulse equivalent data in the first-in-first-out data queue.

[0094] The central processing unit of the local control system calls the moving average filtering algorithm to accumulate and average the extracted queue data. The reconstruction calculation formula for this smoothing process is as follows: ; in, Represents the global dynamic impulse equivalent of the reconstruction operation output; The queue depth parameter represents the total number of data samples participating in the filtering operation; This represents the sorting index number of the data stored in the queue. This indicates the number of data stored in the data queue. The local pulse equivalent value.

[0095] Through the above calculations, the system no longer relies on fixed mechanical transmission ratio parameters. Instead, it extracts and integrates the comprehensive deviation characteristics representing the current real mechanical resistance, belt elastic deformation, and load wear state of the conveyor belt into a smooth global dynamic pulse equivalent. This parameter undergoes high-frequency self-correction iteration as the operating status of the production line changes.

[0096] See attached document Figure 7 After completing the global smooth reconstruction of the pulse equivalent, the system uses the reconstruction parameters to perform feedforward compensation calculations on the trigger coordinates of the underlying control commands.

[0097] The adaptive correction process for the target execution pulse specifically includes the following steps: S306, retrieve the initial state record and baseline parameters of the target object.

[0098] The local control system accesses the execution queue in its internal memory, locates the spatiotemporal joint identifier corresponding to the workpiece at the current trigger position verification sensor, and extracts the initial pulse value latched in the previous step from the data structure of this spatiotemporal joint identifier. Simultaneously, the processor of the local control system calls the second distance data stored during the system initialization phase. This second distance data represents the absolute mechanical span between the photoelectric switch trigger line at the detection station and the center action point of the main rejection actuator.

[0099] S307, Calculate the target execution pulse after slip compensation.

[0100] The local control system performs multiplication-addition hybrid operations based on the extracted initial pulse value, second distance data, and the globally dynamic pulse equivalent generated by previous calculations. The system calculates the theoretical pulse threshold that the spindle absolute encoder should output when the current workpiece reaches the center action point of the main rejection actuator.

[0101] The formula for this calculation is as follows: ; in, This indicates the target execution pulse after adaptive compensation for mechanical slip. This indicates the initial pulse value latched by the hardware when the workpiece is triggered at the inspection station; This represents the second distance data from the inspection station to the main rejection actuator; This represents the global dynamic pulse equivalent output of the system based on the first-in-first-out data queue reconstruction. Through this calculation process, the system maps a constant distance in space into a dynamic pulse increment that includes the current conveyor belt friction resistance, load deformation, and slippage error.

[0102] S308, Update the memory control status of the waiting execution queue.

[0103] The local control system acquires the target execution pulse calculated above and writes it into the designated memory address region of the corresponding workpiece in the waiting execution queue. This write operation overwrites the fixed execution coordinates originally estimated by the system based on the static initial standard pulse equivalent, completing the adaptive correction of the underlying trigger threshold.

[0104] The updated target execution pulse will serve as the sole position comparison benchmark for the hardware driver module to determine whether to issue a solenoid valve trigger action in the subsequent kinematic closed-loop execution logic. For data overwriting and pointer offset operations in the memory address space, those skilled in the art can implement them using conventional structure variable assignment instructions; such memory read / write operations are well-known techniques in the field and will not be elaborated upon here.

[0105] See attached document Figure 8 During the continuous transport process, the local control system processes the logical operation results returned from the remote end through an asynchronous communication mechanism.

[0106] To address potential changes in process standards during production, the system incorporates a specific state machine transition mechanism to prevent control errors caused by sudden parameter changes. The asynchronous instruction reception and system state machine transition process specifically includes the following steps: S401 receives and parses asynchronous network sorting instructions.

[0107] An independent communication thread running within the local control system continuously polls the Ethernet controller's receive buffer in the background to obtain command messages returned by the remote control system. The local control system performs application-layer unpacking of the command message, extracts the sequence number contained in the message header, and uses this as an index to locate the corresponding workpiece record in the local execution queue.

[0108] The specific lower-level characteristics of the aforementioned instruction message are formatted byte stream data carried by Transmission Control Protocol (TCP) or User Datagram Protocol (UDP). For the establishment of Ethernet sockets, cyclic redundancy check (CRC) of data packets, and the logic for reconnection after disconnection, those skilled in the art can implement them by calling standard network communication function libraries. The underlying network communication configuration is well-known technology in this field and will not be elaborated upon here.

[0109] S402, identify the parameter hot update identifier and extract the new detection threshold.

[0110] After parsing the basic sorting instructions, the local control system further reads the parameter status field within a specific offset address of the message. The system determines whether a change has occurred in the characteristic bit within this parameter status field. If a change is detected, the system determines that the remote control system has issued a parameter change instruction. The local control system extracts the new detection threshold from the message's payload data segment. This new detection threshold represents the latest process judgment benchmark that the current production line needs to execute immediately, including new surface defect tolerance limits or new geometric dimensional tolerance ranges.

[0111] S403 executes the switching logic of the system state machine.

[0112] After the local control system acquires the new detection threshold, it interrupts the regular direct memory overwrite operation. The system calls the memory allocation program to allocate an independent address space in the internal random access memory, isolated from the initial global detection threshold. The local control system then writes the new detection threshold into the newly allocated address space.

[0113] Based on the existing state of the system's internal variables, the central processing unit (CPU) of the local control system controls the state machine variables to change. When the message parsing confirms that a new detection threshold has been extracted, the system actively exits the conventional single-buffer operation mode that only maintains a single set of detection benchmarks, and switches to a double-buffer state in which both the old and new sets of valid judgment parameters are stored in memory simultaneously; if the message verification shows that the parameter update field is blank, the system maintains the original single-buffer conventional operation state unchanged.

[0114] Through the aforementioned state machine transition logic, the local control system enters a double-buffer state without interrupting the continuous operation of the mechanical drive device, thus establishing a data processing environment for workpieces that remain on the conveyor belt during the critical period of subsequent parameter changes.

[0115] After the system enters double-buffer mode, the local control system needs to accurately isolate the workpieces in transit that are affected by parameter changes in order to prevent malfunctions caused by sudden changes in process standards.

[0116] The specific process of locking the workpiece space in the gray area and performing secondary logic comparison includes the following steps: The local control system traverses the waiting execution queue in its internal memory, reading the initial pulse value and the target execution pulse updated by the previous steps for each suspended workpiece. Through numerical comparison logic, the system extracts workpieces that meet specific spatial displacement conditions, constructing a gray zone workpiece set. Specifically, the system traverses and compares the memory data, filtering out workpiece records where the current real-time feedback pulse value is greater than or equal to the initial detection pulse, but still less than the target execution pulse. This pulse interval characteristic physically represents that the corresponding workpiece has left the detection origin but has not yet fully arrived at the final sorting execution point.

[0117] The local control system marks all in-transit workpiece records that fall within this range with a gray zone identifier and accumulates the statistical quantity to assign to the total number of gray zone workpieces set in the system. The above locking process makes the system clear about the specific object range that needs to be intervened by double buffering logic.

[0118] S405, extract the local cached raw detection feature data of the gray area workpiece.

[0119] To perform secondary parameter comparison independently locally without needing to request data from the remote control system via the network again, the local control system configures a local feature buffer during the initialization phase. For each locked workpiece within the gray zone workpiece set, the local control system reads the serial number from its spatiotemporal joint identifier. Using this serial number as an index key, the system addresses and retrieves the original feature scalar of that workpiece acquired at the inspection station from the local feature buffer. The lower-level features of these original feature scalars are specifically represented as pixel-level scratch area values ​​output by machine vision algorithms, or millimeter-level outer diameter measurements output by measurement gratings.

[0120] For the memory address mapping mechanism and hash table index retrieval algorithm of local cache, those skilled in the art can use conventional data structures to develop it. The memory lookup algorithm is a well-known technology in this field and will not be described in detail here.

[0121] S406 performs parallel Boolean logic comparisons based on dual thresholds. The central processing unit of the local control system uses the extracted raw feature scalars to call the initial global detection threshold and the new detection threshold stored in memory respectively, and performs parallel arithmetic comparison operations.

[0122] The system determines whether the original feature scalar is within the acceptable range specified by the corresponding threshold. When the feature scalar does not exceed the threshold tolerance range, the system outputs a Boolean value of 1, indicating that the workpiece is qualified; when the feature scalar exceeds the threshold tolerance range, the system outputs a Boolean value of 0, indicating that the workpiece is unqualified.

[0123] Through low-level parallel Boolean comparison operations, the system extracts the same original feature scalar from the local feature cache and compares it with the legal interval boundaries of the old version's initial global detection threshold and the newly issued detection threshold.

[0124] The system thus independently outputs a first decision Boolean value based on the old process benchmark and a second decision Boolean value based on the new process benchmark for a single physical workpiece within the same calculation cycle. This secondary comparison mechanism clearly decouples the legality of the workpiece's state evolution before and after the parameter hot update, providing complete logical support for subsequently issuing accurate sorting channel mapping instructions.

[0125] After completing the secondary logic comparison of the gray zone workpiece based on the new and old detection thresholds, the local control system needs to map the Boolean logic comparison results at the pure software level into channel commands to drive the mechanical equipment.

[0126] The specific process of distributing and isolating the sorting routes includes the following steps: S407 performs routing state space mapping operations.

[0127] The central processing unit of the local control system extracts the old and new logic decision results obtained from parallel computation for the same gray area workpiece. The system internally defines a routing state space containing three independent states, represented as follows: The central processing unit invokes pre-defined multi-condition branching logic to calculate the final execution route for the workpiece.

[0128] The calculation formula for this routing allocation logic is as follows: ; in, This represents the execution routing variable assigned by the system to the current workpiece; Indicates the status of the passageway; Indicates the status of the main rejection channel; Indicates the status of the auxiliary isolation channel; This represents the Boolean value obtained from the initial global detection threshold calculation; This represents the Boolean value obtained based on the new detection threshold calculation.

[0129] S408 establishes the lower-level mapping relationship between sorting routes and hardware drive signals.

[0130] The aforementioned functionally generalized channel states require a strong association between the local control system and specific electrical components at the underlying level. Specifically, when determining... and When all values ​​are equal to 1, it indicates that the parameter change has not altered the workpiece's inherent qualification, and the system allocates a release channel. The underlying logic of this state mapping is to keep the corresponding digital output channel at a low level, preventing the driving of any actuators.

[0131] When the judgment and When all values ​​are equal to 0, it indicates that the parameter change has not altered the unqualified nature of the workpiece. The system allocates the main rejection channel. The underlying logic of this state mapping is to activate the high level of the main drive pin. Its corresponding lower-level characteristic is to drive the solenoid valve bound to the main rejection actuator to engage, thereby controlling the main pusher cylinder to extend.

[0132] When the judgment and When the values ​​are inconsistent, it indicates that the workpiece is a critical workpiece whose state is reversed due to parameter changes. The system allocates an auxiliary isolation channel. The underlying logic of this state mapping is to activate a high level on the auxiliary drive pin. The corresponding lower-level characteristic is to drive the solenoid valve bound to the auxiliary isolation actuator to engage, thereby controlling the auxiliary diversion plate cylinder or the falling flap action. This isolation action physically separates the critical workpiece from normal products and absolute scrap, allowing it to flow into the manual review buffer.

[0133] S409, write the route execution flag and release the workpiece from the pending state.

[0134] After calculating the execution route, the local control system addresses the corresponding memory block in the execution queue based on the sequence number in the spatiotemporal joint identifier of the workpiece. The system then overwrites the execution route into the route identifier register within that memory block. Subsequently, the local control system clears the suspended status flag of the workpiece in the execution queue, changing it to an execution-allowed state. At this point, the workpiece within the gray area, subject to double-buffered logic intervention, possesses a complete trigger target and a clear action direction, awaiting scanning and triggering by the underlying kinematic closed-loop program.

[0135] For bitwise logic operations and interrupt masking of the underlying memory block registers, those skilled in the art can implement them using conventional microcontroller register operation instructions. The memory set and reset operations are well-known technologies in the field and will not be described in detail here.

[0136] After the system completes parameter reconstruction and logical routing allocation, it needs to translate the instructions in the data space into precise spatial actions. The local control system relies on underlying real-time tasks to continuously monitor the spatial coordinate status of the conveying equipment to ensure that the actuators intervene at the accurate location.

[0137] The specific process of real-time pulse closed-loop comparison and trigger threshold determination includes the following steps: S501 performs low-level high-frequency polling and reads the current spatial feedback coordinates.

[0138] The low-level scanning task of the local control system runs continuously in the background at a set fixed control cycle. Within each scanning cycle, the processor calls the hardware register read interface of the high-speed counting module to obtain the real-time value currently output by the spindle absolute encoder. The local control system then assigns this value to the current pulse variable.

[0139] The specific underlying characteristic of the aforementioned low-level scanning task is a periodic interrupt routine triggered by the microcontroller hardware timer within the local control system. The setting of the trigger frequency and priority configuration of the timer hardware interrupt can be implemented by conventional register initialization code programming by those skilled in the art; the underlying interrupt configuration method is well-known in the field and will not be elaborated upon here.

[0140] S502, access the waiting execution queue and retrieve the workpiece data to be compared.

[0141] After reading the current pulse variable, the local control system accesses the waiting execution queue in the internal memory through the memory bus. According to the first-in-first-out address sorting principle of the data queue, the system reads the workpiece data structure at the front of the queue with the status flag in the allowed execution state. The system parses the contents of the structure and extracts the target execution pulse and sorting route determined in the previous calculation for the specific workpiece.

[0142] S503 performs closed-loop comparison logic operations on absolute pulse coordinates.

[0143] The arithmetic logic unit of the local control system uses the current pulse variable read in real time as input parameter and the target execution pulse extracted from the queue as boundary comparison benchmark to determine the difference in the hardware action trigger condition. If the current feedback pulse value has not yet reached the set target benchmark, it indicates that the workpiece is in the intermediate transmission segment between the position verification sensor and the actuator. The system control hardware action trigger condition flag bit remains in the reset state, and no action is executed.

[0144] If the current feedback pulse value has reached or exceeded the target reference limit, it indicates that the cumulative running displacement of the conveyor belt has met the compensation conditions, the centroid of the workpiece has accurately coincided with the mechanical execution reference point, and the system immediately activates the hardware action trigger condition flag.

[0145] S504 generates underlying hardware trigger events.

[0146] Within the control cycle when the hardware action trigger condition flag bit of the difference judgment calculation output is equal to 1, the local control system stops the position monitoring polling task of the workpiece, extracts the sorting route attached to the workpiece data structure, uses the sorting route as the action parameter, and generates the underlying hardware trigger event in combination with the system clock of the current control cycle.

[0147] The system pushes the underlying hardware trigger event into the action execution buffer to provide a basis for the electrical port drive in the next stage. In this process, the system relies on the comparison of absolute pulses to form closed-loop control, avoiding the delay trigger error caused by the speed adjustment of the conveyor belt.

[0148] After the underlying hardware trigger event is generated, the local control system needs to convert the sorting route at the logic level into a high-voltage control signal at the physical level. The local control system drives the specific mechanical actuator through the digital output port to complete the separation of the workpiece.

[0149] The specific process of multi-channel action distribution and hardware driver includes the following steps: S505 parses underlying hardware trigger events and maps output port status.

[0150] The local control system reads the underlying hardware trigger events in the action execution buffer, extracts the sorting routes bound to them, and the internal processor of the local control system directly maps and calculates the logic level state combination of the main and secondary pins of the underlying digital output module based on the parsed sorting routes.

[0151] When the sorting route result points to the release channel, the system controls the electrical pins corresponding to the main rejection and auxiliary isolation to maintain a low-level, non-driven resting state.

[0152] When the sorting routing result points to the main rejection channel, the system is configured to output a high-level signal on the pin connected to the main rejection actuator.

[0153] When the sorting routing result points to the auxiliary isolation channel, the system is configured to separately configure the pin connected to the auxiliary isolation diversion device to output a high level to trigger the action.

[0154] S506 issues electrical drive commands and executes sorting actions.

[0155] Based on the calculated output port state vector, the local control system performs set or reset operations on the corresponding digital output module pins.

[0156] When the sorting route is in the main rejection channel state, the local control system sets the main pin to a high level. This main pin is connected to the main rejection actuator. This high-level signal drives the solenoid valve coil attached to the main rejection actuator to energize and engage. Compressed air enters the rodless chamber of the main pusher cylinder, pushing the piston rod of the main pusher cylinder to extend and push the defective workpieces on the conveyor belt into the waste collection box.

[0157] When the sorting route is in auxiliary isolation channel mode, the local control system sets the auxiliary pin to a high level. This auxiliary pin is connected to the auxiliary isolation actuator, and the high-level signal drives the solenoid valve coil attached to the auxiliary isolation actuator to energize and engage. Compressed air enters the rodless chamber of the auxiliary diverter cylinder, pushing the diverter to swing at a specific angle, guiding the workpieces on the conveyor belt that are at the critical state of parameter change to the manual verification buffer.

[0158] When the sorting route is in the release channel state, the local control system keeps the main pin and auxiliary pin in a low level state, the main rejection actuator and the auxiliary isolation actuator remain in place, and the workpiece flows smoothly into the next qualified product packaging process with the conveyor belt.

[0159] For the pin drive circuit design of digital output modules, the relay isolation amplifier circuit of solenoid valve coils, and the mechanical anti-interference design of cylinder action, those skilled in the art can consult conventional mechatronics equipment manuals for selection and configuration. Their electrical drive and pneumatic control circuits are well-known technologies in this field and will not be described in detail here.

[0160] S507 controls the output pulse width and resets the actuator.

[0161] While the output pin is set to high, the local control system starts an internal timer, loading a preset action hold time parameter. This action hold time parameter represents the theoretical time required for the cylinder to complete its full extension stroke. Once the timer reaches the set value, the local control system generates a hardware interrupt. The processor responds to this interrupt by clearing the corresponding digital output pin's level to zero.

[0162] When the solenoid valve coil is de-energized, the cylinder returns to its initial position under the action of the internal return spring or reverse air pressure, completing the mechanical loop of a single control cycle. This process prepares the machine mechanically for the next workpiece sorting operation.

[0163] After the actuator resets and completes the sorting operation, the local control system needs to reclaim the memory resources allocated in the internal random access memory. This process ensures that the memory space is not exhausted during continuous high-frequency operation and guides the system to smoothly exit the double-buffered state caused by parameter changes.

[0164] The specific process of lifecycle management and memory destruction includes the following steps: S508, release the workpiece data memory and perform a dequeue operation.

[0165] After confirming that the digital output pin level is cleared, the local control system accesses the waiting execution queue in the internal memory. According to the first-in-first-out address sorting principle of the data queue, the system pops the data structure corresponding to the workpiece that has completed the sorting action from the head of the queue. The system calls the memory release interface to destroy the memory block occupied by the spatiotemporal joint identifier of the workpiece.

[0166] The system extracts the serial number of the workpiece and uses it as a memory addressing index to locate the address block containing the original feature scalar corresponding to the workpiece in the local feature cache. The system clears the values ​​in this address block to zero and marks its memory space as available and free. Through this memory destruction mechanism, the system completes the full lifecycle data tracking and cleanup of a single workpiece from the moment it is triggered to the completion of its action.

[0167] For the memory release logic of the underlying data structure and the reclamation of pointer address space, those skilled in the art can use conventional microcontroller dynamic memory management libraries for programming implementation. The memory defragmentation and reclamation mechanism is a well-known technology in this field and will not be described in detail here.

[0168] S509, update the total number of gray zone workpieces and perform status determination.

[0169] When processing gray zone workpieces that are at the critical point of parameter change, the local control system needs to synchronously maintain system-level status indicator variables, and the system reads the total number of gray zone workpieces stored in the internal register. Whenever a workpiece with a gray zone identifier completes the above dequeue and memory destruction operations, the central processing unit decrements the total number of gray zone workpieces variable.

[0170] After decrementing the total number of gray zone workpieces stored internally, the system immediately compares the updated remaining value with the zero value for verification. If the comparison shows that the total number is still greater than zero after deduction, it proves that there are still critical workpieces on the conveyor belt that have not yet reached the actuator and are affected by parameter changes. The system controls the gray zone clearing flag to maintain a negative judgment and continues the double-buffered isolation intervention mode. If the result shows that the total number after decrementing is completely zero, it proves that all in-transit workpieces within the gray zone locking range have completed the sorting and memory destruction process. The system then activates the gray zone clearing flag to confirm that the data boundary conditions for exiting the special intervention mode are met.

[0171] S510 triggers system state machine rollback and memory pointer redirection.

[0172] Within the same control cycle in which the gray zone clear flag is detected to be activated, the local control system immediately initiates a global rollback intervention for the entire process. Based on the clear verification command, the system's central processing unit forcibly cuts off the maintenance logic of the double-buffered state, and hard switches the currently running control state machine variables from the multi-judgment standard coexistence mode back to the single-buffered normal operation state that only refers to a single valid standard.

[0173] After the state machine successfully switches back to the regular single-buffered working surface, the local control system initiates the redirection configuration of the underlying memory pointer. The system forcibly offsets the pointer designated by the system for global reference parameter addressing from the old memory address where the original initial global detection threshold is located and points to the independent address space where the new detection threshold is located. The system then overwrites the specific value of the new detection threshold into the system's globally unique reference decision register.

[0174] After overwriting the benchmark decision register, the system calls a memory release instruction to invalidate the old memory space occupied by the initial global detection threshold. At this point, the remote control system, through asynchronous network messages, officially takes over the logical decision-making of all newly entered workpieces at the inspection station using the new process parameters. The system achieves a seamless closed-loop handover between the control benchmark and the underlying mapping logic without interrupting the continuous operation of the equipment drive.

[0175] Specific application examples: Basic environment and parameter settings: This system is deployed on a continuous production line for U-shaped hanging ring iron accessories. The local control system loads parameters during the initialization phase. The first distance data is input as 1000mm (distance from the inspection station to the position verification sensor), and the second distance data is input as 2500mm (distance from the inspection station to the main rejection actuator). The initial standard pulse equivalent of the spindle absolute encoder is set to 10.00 pulses / mm. The initial global detection threshold is set as follows: surface scratch area less than or equal to 10mm². 2 (Qualified).

[0176] Pulse reconstruction and mechanical slip compensation: The current processing batch has entered continuous operation. The target workpiece enters the inspection station and blocks the photoelectric switch. The local control system latches the initial pulse value of 50000 and extracts the system timestamp to assemble the spatiotemporal joint identifier.

[0177] When the target workpiece moves to the position verification sensor, it is blocked and triggered. The local control system latches the real pulse value as 60050.

[0178] The system calls the formula to perform local impulse equivalent calculation: =(60050−50000) / 1000=10.05 (pulses / mm).

[0179] The system pushes 10.05 into the FIFO data queue. Assuming the previous historical values ​​in the queue are [10.04, 10.06, 10.05, 10.04], the system uses a filtering algorithm to accumulate and average these values, outputting the global dynamic pulse equivalent. =10.048.

[0180] System execution target pulse compensation calculation: =50000+2500×10.048=75120.

[0181] Note: Without a compensation mechanism, the theoretical pulse count would be 50000 + 2500 × 10.00 = 75000. The dynamic compensation calculation smooths out the actual 120 pulse slip errors that occurred on the conveyor belt.

[0182] Double-buffered state switching and logical isolation: While the target workpiece is en route, the remote server sends a new process standard message, requiring a tighter tolerance for scratch area. The local control system then extracts a new detection threshold: surface scratch area less than or equal to 5 mm². 2 .

[0183] The local control system switches to double-buffered mode and scans the execution queue. At this time, the real-time feedback pulse is 65000, which is greater than the initial pulse value (50000) and less than the target execution pulse (75120). The system locks the current workpiece as a gray zone workpiece.

[0184] The system retrieves the original image detection results of the current workpiece from the local feature cache, showing that the actual scratch area is 8mm. 2 The system performs a dual-track Boolean logic comparison: Compare with the initial global detection threshold (tolerance 10): =1; Compare with the new detection threshold (tolerance 5): =0.

[0185] because The calculation result triggers an independent decision branch, and the local control system writes the auxiliary isolation channel identifier into memory.

[0186] Closed-loop execution: When the underlying polling monitoring detects that the real-time pulse of the spindle absolute encoder has reached 75120, the local control system activates the digital output pin of the associated auxiliary isolation actuator, energizes the solenoid valve, and activates the flow divider, pushing the critical workpiece causing process standard disputes into the manual review buffer. After the action is completed, the system releases the corresponding memory space. Once all gray zone workpieces have been processed, the system automatically overwrites the initial parameters and resumes normal single-buffer operation.

[0187] Experimental verification and effect comparison: Experimental Verification 1: A 24-hour uninterrupted operation test was conducted on a standard U-shaped hanging ring iron accessory continuous production line. The conveyor belt was set to the rated maximum operating speed, and different weight loads were randomly added during the operation to simulate the load changes in the real workshop.

[0188] The control group used conventional static pulse control logic, which calculates coordinates using a constant pulse equivalent parameter set during initialization. The test group used the control logic proposed in this invention, which reconstructs the global dynamic pulse equivalent based on a position verification sensor and a first-in-first-out data queue.

[0189] Core data collection metrics: At the set running time nodes, the actual coordinates of the workpiece centroid reaching the main rejection actuator are captured by an external high-precision laser rangefinder. The position deviation value between the system command trigger and the absolute zero point is recorded, and the number of times the solenoid valve or cylinder is stuck due to position triggering error is counted. The specific test data is shown in Table 1.

[0190] Table 1. Comparison data of trigger coordinate deviation under continuous operation. in conclusion: Observe the data in Table 1 and combine them with... Figure 9 The position accuracy comparison curves of the control schemes show that, with the extension of equipment operating time, the position accuracy curve of the control group exhibits a divergent characteristic of linear drift and periodic fluctuations due to belt wear, thermal expansion and contraction, and load slip. The position deviation shows a significant divergence trend, with the maximum deviation exceeding 16mm, resulting in a high mechanical jamming failure rate. These phenomena indicate that static logic cannot eliminate the cumulative systematic errors introduced by belt deformation and mechanical wear.

[0191] This invention achieves adaptive absorption and compensation for slip variables by relying on local pulse equivalent extraction and global dynamic smoothing calculation. In the comparison curves, the error sampling points of the control scheme of this invention consistently converge stably near the zero baseline, without any drift phenomenon accumulating over time. Throughout the entire 24-hour operating cycle, the position deviation is consistently kept within a small tolerance range of ±0.8mm, ensuring the execution accuracy for long-term continuous operation.

[0192] Experimental Verification 2: During the 12th hour of continuous system operation, a new testing and judgment formula was suddenly issued via the remote control system. The new process requires that the upper limit of the allowable area of ​​scratches on the surface of iron accessories be increased from the initial 10mm. 2 Tighten to 5mm 2 .

[0193] At the moment this network message was sent and took effect, 50 workpieces with completed image acquisition were stuck on the conveyor belt in the area downstream of the inspection station but not yet reaching the actuator. After offline verification, 8 of these 50 workpieces had actual scratches with an area between 5mm and 6mm. 2 Up to 10mm 2 The critical workpiece between [the specified points]. For this sudden working condition, the intervention results of three different coping strategies were recorded. Specific test data are shown in Table 2.

[0194] Table 2. Comparison of Results of Hot-Update Processing of Detection Thresholds under Different Control Strategies in conclusion: As can be seen from the data in Table 2, although the conventional shutdown and emptying strategy did not produce incorrect sorting, it required interrupting the drive for 15 minutes for manual cleaning of residual materials, which disrupted the continuous operation cycle of the equipment and reduced the overall efficiency of the equipment.

[0195] The conventional direct memory overwrite strategy forcibly overwrites the old threshold with the new threshold without stopping the system. This causes the above 8 critical workpieces to lose their accurate benchmark during the transition between the old and new logic. All of them are incorrectly classified and mixed into the final qualified product packaging box, resulting in batch contamination.

[0196] The system of this invention relies on constructing a dual-buffer state to accurately lock 50 gray-zone objects in transit, and independently completes dual-Boolean parallel logic decoupling for 8 critical workpieces in the background. Without interrupting the production line at any time, this system flawlessly assigns critical workpieces prone to quality disputes to auxiliary isolation channels for separation actions. This experimental data confirms the effectiveness of this invention in resolving technical contradictions arising from dynamic changes in process parameters.

Claims

1. A remote control system for a finished iron accessory inspection and sorting machine, characterized in that, include: The finished iron accessory inspection and sorting machine is used to carry and transport the workpieces to be tested, provide trigger signals and perform sorting actions. The finished iron accessory inspection and sorting machine includes a conveyor belt, a spindle absolute encoder, an inspection station, a position verification sensor, a main rejection actuator and an auxiliary isolation actuator. A local control system is used to connect to the finished iron accessory inspection and sorting machine. The local control system includes a data acquisition module, a pulse reconstruction module, and a routing execution module. A remote control system for communicating with the local control system, the remote control system including a command issuing and parameter management module; The data acquisition module is used to generate a spatiotemporal joint identifier for the workpiece under test based on the output of the detection station and the spindle absolute encoder, and upload the detection data to the remote control system. The instruction issuance and parameter management module is used to generate sorting instructions based on the detection threshold and return them to the local control system. The pulse reconstruction module is used to calculate and update the target execution pulse of the workpiece under test by combining the trigger state of the position verification sensor; The routing execution module is used to allocate sorting routes and output electrical drive signals to the main rejection execution mechanism or the auxiliary isolation execution mechanism.

2. The remote control system for a finished iron accessory inspection and sorting machine according to claim 1, characterized in that, The detection station is equipped with a trigger photoelectric switch and an image acquisition device. When the workpiece under test enters the detection station and blocks the trigger photoelectric switch, the data acquisition module extracts the system timestamp, latches the initial pulse value output by the spindle absolute encoder, and generates a serial number. The data acquisition module combines the sequence number, the system timestamp, and the initial pulse value to construct the spatiotemporal joint identifier; The data acquisition module encapsulates the detection data obtained by the image acquisition device and the spatiotemporal joint identifier into a network message and uploads it to the remote control system.

3. The remote control system for a finished iron accessory inspection and sorting machine according to claim 2, characterized in that, The position verification sensor is located downstream of the detection station; When the workpiece under test triggers the position verification sensor, the pulse reconstruction module extracts the actual pulse value at the trigger moment; The pulse reconstruction module extracts the initial pulse value from the spatiotemporal joint identifier; The local control system reads the first distance data from the detection station to the position verification sensor; The pulse reconstruction module combines the actual pulse value, the initial pulse value, and the first distance data to calculate and obtain the local pulse equivalent.

4. The remote control system for a finished iron accessory inspection and sorting machine according to claim 3, characterized in that, The pulse reconstruction module has an internal first-in-first-out data queue. The pulse reconstruction module determines the address usage status of the first-in-first-out data queue; If the first-in-first-out data queue is not full, the pulse reconstruction module appends the local pulse equivalent to the tail of the first-in-first-out data queue. The pulse reconstruction module extracts all existing historical local pulse equivalent data in the first-in-first-out data queue, performs filtering operations, and generates a global dynamic pulse equivalent.

5. The remote control system for a finished iron accessory inspection and sorting machine according to claim 4, characterized in that, The local control system reads the second distance data from the detection station to the main rejection actuator; The pulse reconstruction module uses the global dynamic pulse equivalent and the second distance data, combined with the initial pulse value, to perform adaptive compensation calculation; The pulse reconstruction module calculates and updates the target execution pulse corresponding to the workpiece under test through the adaptive compensation calculation. The local control system writes the target execution pulse into the waiting execution queue.

6. The remote control system for a finished iron accessory inspection and sorting machine according to claim 5, characterized in that, The instruction issuance and parameter management module receives the network packet, performs logical operations on the detection data according to the currently effective detection threshold, and generates the sorting instruction containing the judgment result and the spatiotemporal joint identifier; When the instruction issuance and parameter management module issues the sorting instruction containing the new detection threshold, the local control system allocates an independent address space to write the new detection threshold. The routing execution module switches to double-buffered mode.

7. The remote control system for a finished iron accessory inspection and sorting machine according to claim 6, characterized in that, After the routing execution module switches to the double-buffered state, the local control system reads the initial pulse value and the target execution pulse corresponding to the in-transit workpiece in the waiting execution queue; The local control system continuously reads the current pulse value of the spindle absolute encoder and extracts the in-transit workpieces whose current pulse value is greater than or equal to the initial pulse value and less than the target execution pulse to construct a gray zone workpiece set. The routing execution module locks the gray zone workpieces within the gray zone workpiece set.

8. The remote control system for a finished iron accessory inspection and sorting machine according to claim 7, characterized in that, The local control system is set with an initial global detection threshold and pre-stores the original feature scalars of the workpiece being tested. The routing execution module extracts the original feature scalar corresponding to the gray area workpiece; The routing execution module performs a secondary logical comparison between the original feature scalar and the initial global detection threshold and the new detection threshold, respectively. The routing execution module independently outputs a first decision Boolean value based on the initial global detection threshold and a second decision Boolean value based on the new detection threshold for a single gray area workpiece.

9. The remote control system for a finished iron accessory inspection and sorting machine according to claim 8, characterized in that, The routing execution module assigns the corresponding sorting route to the gray zone workpiece based on the first determination boolean value and the second determination boolean value. The sorting route includes a release channel, a main rejection channel, and an auxiliary isolation channel; If the first and second Boolean values ​​are inconsistent, the routing execution module allocates the auxiliary isolation channel to the gray zone workpiece.

10. The remote control system for a finished iron accessory inspection and sorting machine according to claim 9, characterized in that, The local control system continuously reads the current pulse value of the spindle absolute encoder; When the current pulse value is greater than or equal to the target execution pulse, the routing execution module outputs the electrical drive signal to the main rejection execution mechanism or the auxiliary isolation execution mechanism to perform the corresponding action according to the assigned sorting route; After the corresponding action is completed, the local control system releases the memory space occupied by the gray area workpiece, and deletes the initial global detection threshold and exits the double buffer state after the number of gray area workpieces is reduced to zero.

Images