Distributed signal transmission and linkage control system for automated warehouses and material unloading lines

By dynamically adjusting channel priority and generating secondary arbitration tokens between the automated warehouse and the unloading line nodes, combined with segmented hash verification and forward redundant transmission, the communication latency and reliability issues in the distributed linkage control system are solved, and efficient linkage control is achieved.

CN122698935BActive Publication Date: 2026-11-17AVIC SPECIAL MATERIALS IND (XIAN) CO LTD +1
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
CN202611201556.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2026-08-10
Publication Date
2026-11-17
Estimated Expiration
2046-08-10

AI Technical Summary

Technical Problem

Existing distributed linkage control systems suffer from problems such as static priority, inefficient conflict handling, rigid transmission mechanisms, and poor linkage reliability in communication between automated warehouse and material unloading line nodes. This leads to increased network latency and fails to meet the requirements of high real-time linkage control.

Method used

By deploying a primary channel and multiple backup channels between the automated warehouse and the unloading line nodes, a weighted score is calculated based on the historical bit error rate, occupancy time, and distance of the backup channels. The priority sequence is dynamically adjusted, and a secondary arbitration token is generated using MAC address and timestamp to resolve network conflicts. Combined with segmented hash verification and forward redundant transmission mechanism, the transmission reliability is improved.

Benefits of technology

It enables real-time transmission of emergency tasks, reduces production line downtime and collaborative operation delays, improves the fairness of network resource allocation and data integrity, and enhances the reliability of the linkage between the automated warehouse and the material unloading line.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122698935B_ABST
    Figure CN122698935B_ABST
Patent Text Reader

Abstract

This invention belongs to the field of control technology, specifically relating to a distributed signal transmission and linkage control system for automated storage and retrieval systems (AS / RS) and material handling lines. It includes a selection module for deploying a main channel and multiple backup channels between the AS / RS and material handling line nodes. The module calculates a weighted score based on the historical bit error rate, occupancy time, and distance of the backup channels, and sorts them in descending order of score to obtain an initial priority sequence. When the main channel experiences a data interruption, priority arbitration is triggered. The weighted score of each backup channel is multiplied by a correction factor that is positively correlated with task urgency and negatively correlated with the offset of the corresponding storage location. The backup channel with the highest correction score is selected for switching. This invention improves the reliability of the linkage and coordination between the AS / RS and material handling lines by bundling and slicing control signals, combined with a forward redundant transmission strategy based on channel ranking and a retransmission over-limit attenuation and reordering mechanism, thereby reducing production line downtime and collaborative operation delays.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of control technology, specifically relating to a distributed signal transmission and linkage control system for automated warehouses and material unloading lines. Background Technology

[0002] In a distributed architecture, the automated storage and retrieval system (AS / RS) and each node of the unloading line need to exchange equipment status and control commands. Because there is a time-series binding between the AS / RS's outbound actions and the unloading line's start and stop, any delay or loss of data transmission will cause a disconnect between the two. Once communication is interrupted, triggering a linkage failure, it will cause intermittent material conveying, accumulation and congestion, and even mechanical interference and equipment damage, threatening the safety and continuous operation of the production line.

[0003] Chinese patent document CN109639546B discloses a redundant system based on CAN bus. This patent can effectively overcome the defects of dual redundant buses in master-slave mode, such as the inability to mix and mount, master-slave switching disturbance, and the use of bit fields in the arbitration or data domain for redundant frame arbitration in mutual backup mode. It realizes that redundant frame identification can be performed without waiting, thereby greatly improving the real-time performance of the system. It also enables the use of third-party application layer protocols without modification, avoiding the problem of only being able to run custom protocols and expanding the scope of application of redundant protocols.

[0004] Existing distributed linkage control systems generally utilize channel redundancy and priority arbitration techniques, deploying a main channel and multiple backup channels between the automated storage and retrieval system (AS / RS) and unloading line nodes. When the main communication link fails, priorities are assigned based on simple channel bandwidth or fixed network topology parameters, switching to backup channels to restore network connectivity. Data integrity is ensured through traditional message retransmission mechanisms in case of data loss. However, priority assessment is often a static mechanism, failing to integrate historical network conditions with the specific business characteristics of the AS / RS, resulting in channel switching failing to adapt to actual operational needs. When multiple nodes access simultaneously, causing arbitration conflicts, existing mechanisms typically employ fixed backoff algorithms, lacking token contention and weight adjustment capabilities, easily exacerbating network congestion. In the message transmission stage, fixed segmentation verification and retransmission mechanisms are not adjusted according to channel quality. When facing link quality degradation, they lack forward redundancy compensation and degradation circuit breaking mechanisms for inferior channels, leading to increased linkage message transmission latency and failing to meet the high real-time linkage control requirements of the AS / RS and unloading line. Summary of the Invention

[0005] This invention provides a distributed signal transmission and linkage control system for automated warehouses and unloading lines to solve the technical problems of static priority, inefficient conflict handling, rigid transmission mechanism, and poor linkage reliability in the prior art.

[0006] This invention provides a distributed signal transmission and linkage control system for automated warehouses and unloading lines, comprising: The selection module is used to deploy a main channel and multiple backup channels between the automated warehouse and the unloading line node. A weighted score is calculated based on the historical bit error rate, occupancy time and distance of the backup channels, and the scores are sorted in descending order to obtain an initial priority sequence. When the main channel experiences a data interruption, priority arbitration is triggered. The weighted score of each backup channel is multiplied by a correction factor that is positively correlated with the urgency of the task and negatively correlated with the offset of the corresponding storage location of the backup channel. The backup channel with the highest correction score is selected for switching. The verification module is used to use the XOR value of the last bit of the node's MAC address and the timestamp as a secondary arbitration token when an arbitration conflict occurs. The node with the smaller XOR value is given priority to access the network, and the Hamming weight of the XOR value is used to adjust the weight of the next round of occupancy time. After the backup channel is switched, the transmitted message is segmented and hashed for verification. The segment length is directly proportional to the correction score and inversely proportional to the offset of the corresponding cargo location of the backup channel. The control module is used to bundle the outbound signal and the material feeding start / stop signal into a linkage message and divide it into sub-messages. The sub-messages are numbered in ascending order according to the data slice sequence. The receiving end requests retransmission based on the missing sub-message number and performs forward redundant transmission of the preceding sub-messages. The number of redundant copies increases as the initial sequence position increases. If the number of retransmissions exceeds the set threshold, the weighted score is decayed to the last position. In the next cycle, the error rate weight is temporarily increased and the scores are reordered to execute linkage control.

[0007] A distributed linkage control system integrating communication quality, spatial dimension, and business attributes was constructed. Through multi-factor weighted initial evaluation and channel switching mechanism with dynamic correction of tasks and storage locations, the real-time transmission of emergency tasks is guaranteed. A two-level token with MAC and timestamp XOR and Hamming weight adjustment is adopted to resolve network conflicts and maintain the fairness of resource allocation. Combined with dynamic segmented hash verification, forward redundant transmission, and retransmission over-limit degradation mechanism, the reliability of linkage between the automated warehouse and the unloading line is improved, effectively reducing production line downtime and coordination delay.

[0008] Furthermore, a weighted score is calculated based on the historical bit error rate, occupancy time, and distance of the backup channel, including: the weighted score of the backup channel is calculated using the following formula: W=Ma×Eb×Tc×D, where W is the dimensionless value of the weighted score, M is the preset dimensionless constant of the basic full score, E is the value of the historical bit error rate of the channel, a is the first preset conversion coefficient, T is the value of the total occupancy time of the channel in milliseconds, b is the second preset conversion coefficient, D is the value of the three-dimensional straight-line distance of the node, and c is the third preset conversion coefficient.

[0009] By using a weighted scoring formula based on historical bit error rate, occupancy time, and spatial distance, the quality of backup channels can be quantified and comprehensively evaluated. Communication stability, channel load, and physical spatial distance are incorporated into a unified evaluation dimension, resulting in high scoring differentiation. This provides an accurate and objective quantitative basis for initial priority ranking, avoids the one-sidedness of evaluation based on a single indicator, and improves the rationality of initial channel screening.

[0010] Furthermore, the weighted score of each backup channel is multiplied by a correction factor that is positively correlated with the urgency of the task and negatively correlated with the offset of the corresponding storage location of the backup channel. The backup channel with the highest correction score is selected for switching. This process includes: extracting a preset integer value representing the urgency level of the task from the master task queue; dividing the preset integer value by the sum of the measured straight-line distance from the automated racking point to the backup channel reference point and a preset zero-prevention constant to obtain a dimensionless quotient; setting the dimensionless quotient as the correction factor for the backup channel; multiplying the weighted score of the backup channel by the correction factor to obtain the correction score; sorting all backup channels from largest to smallest correction score to output a priority sequence; and selecting the backup channel with the highest ranking for switching.

[0011] By introducing a correction factor based on task urgency and cargo location offset, dynamic adaptation of backup channel priority is achieved. Urgent tasks can obtain higher switching priority, while avoiding transmission risks caused by excessive cargo location offset. This allows channel switching to fit the actual warehousing operation needs, significantly improving the transmission timeliness and adaptability of high-urgency linkage tasks.

[0012] Furthermore, the XOR value of the last bit of the node's MAC address and the timestamp is used as the secondary arbitration token. This involves: using the underlying network protocol to read the last 8 bits of the MAC address of the sending node's hardware network card, extracting the last 8 bits of the current millisecond-level clock timestamp variable of the sending device, aligning the two, performing a bit-by-bit binary logical XOR operation through processor instructions, and concatenating the calculated results of the 8 binary bits to form a whole byte binary value, which is then sent to the bus as the secondary arbitration token to perform arbitration allocation.

[0013] By XORing the last bit of the MAC address with a timestamp to generate a secondary arbitration token, the inherent attributes of the device are combined with dynamic time information. The token has strong randomness and high conflict differentiation, which can quickly determine the access priority of conflicting nodes, realize microsecond-level conflict arbitration, avoid the congestion aggravation problem of fixed backoff algorithms, and ensure the high efficiency of network access.

[0014] Furthermore, the segment length is adjusted proportionally to the correction score and inversely proportional to the offset of the corresponding cargo location on the backup channel. This includes: pre-setting the basic message segmentation constant parameter to 2048 bytes in the controller memory; determining the fixed byte length of the single segment packet payload of the channel transmission message; the fixed byte length of the single segment packet payload is the maximum value between the preset positive integer minimum byte constant and the calculated value rounded down; the calculated value is the quotient obtained by dividing the dimensionless correction score of the backup channel by the distance compensation sum and multiplying it by the basic message segmentation constant parameter; the distance compensation sum is the sum of the dimensionless value corresponding to the measured offset distance of the cargo location coordinate axis in meters on the backup channel and the preset dimensionless positive compensation constant.

[0015] A dynamic correlation mechanism is established between message segment length and correction score and cargo location offset. Long segments are used for high-score, near-cargo-location channels to reduce verification overhead, while short segments are used for low-score, far-cargo-location channels to reduce the probability of transmission errors. At the same time, a minimum length is set to avoid fragmentation, balancing transmission efficiency and data integrity, and adapting to different channel quality and spatial scenarios.

[0016] Furthermore, the outbound signal and the material feeding start / stop signal are bundled into a linkage message and divided into sub-messages. The sub-messages are numbered from smallest to largest according to the data slice order, including: The multi-axis motion trajectory data stream of the three-dimensional container outbound action issued by the controller is extracted and written into the assembly buffer to represent the outbound signal. The continuous high-frequency sampled braking potential value sequence representing the motor and brake status of the downstream production line is extracted, filled into the buffer and appended after the data segment to complete the binding and encapsulation of the linkage message whose total data volume exceeds the single frame transmission limit. Extract the natural number 1 as the initial base number. When grouping the linked message into sub-messages, increment the natural number 1 step by step according to the data slice order to obtain a monotonically continuous integer and write it into the serial number slot of the first two bytes of the corresponding sub-message header as the number.

[0017] Furthermore, the preceding sub-messages are sent with forward redundancy, and the number of redundancies increases as the initial sequence position progresses, including: Retrieve the ranking index value of the currently used backup channel in the initial priority sequence, and calculate the total number of failed links of the backup channel that has failed in the current network link pool; The redundancy packet quota is calculated by adding the total number of faulty links to the ranking index value and then adding 1 to this sum. This quota instructs the network card to initiate the corresponding number of synchronous redundant transmissions of the preceding sub-packets.

[0018] Furthermore, Gigabit Ethernet is used as the primary channel, and multiple industrial wireless LANs and fieldbuses of different frequency bands are used as backup channels.

[0019] Furthermore, when the automated warehouse side node and the unloading line side node simultaneously send association requests to the same wireless access point, causing the backoff timer of the carrier sense multiple access conflict avoidance mechanism to be fully loaded, an arbitration conflict is triggered.

[0020] Furthermore, the Brian Kernighan algorithm is used to iteratively remove the least significant integer 1 from the binary representation of the XOR value, and the total number of iterations is counted to obtain the Hamming weight value.

[0021] The beneficial effects are as follows: This invention constructs a linkage control mechanism that integrates communication quality and spatial dimension. It performs an initial evaluation of backup channels by comprehensively considering historical bit error rates, occupancy duration, and distance. In the event of an interruption, it uses task urgency and location offset to correct the score, combining lower-level communication with upper-level warehousing operations to ensure the real-time transmission of urgent tasks. In the face of network conflicts, it uses the XOR value of the MAC address and timestamp as a secondary token and updates the weight for the next round based on Hamming weight, ensuring long-term fairness in resource allocation. Simultaneously, it adjusts the segment length of hash verification based on the corrected score and location offset, reducing communication verification overhead while ensuring data integrity. By bundling and slicing control signals, combined with a forward redundant transmission strategy based on channel ranking and a retransmission over-limit attenuation and reordering mechanism, it improves the reliability of the linkage and coordination between the automated warehouse and the unloading line, reducing production line downtime and collaborative operation delays. Attached Figure Description

[0022] Figure 1 This is a structural block diagram of the distributed signal transmission and linkage control system for automated warehouses and material unloading lines.

[0023] Figure 2 The graph shows the weighted score of the backup channel as a function of the historical bit error rate.

[0024] Figure 3 This is a schematic diagram of the curve showing the change in the load length of a single package as a function of offset distance.

[0025] Figure 4 This diagram illustrates the comparison of performance indicators of different ablation methods in experimental experiments. Detailed Implementation

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

[0027] An embodiment of the distributed signal transmission and linkage control system for automated warehouses and unloading lines provided by this invention: like Figure 1As shown, the distributed signal transmission and linkage control system for the automated warehouse and unloading line includes: The selection module is used to deploy the main channel and multiple backup channels between the automated warehouse and the unloading line node. The weighted score is calculated based on the historical bit error rate, occupancy time and distance of the backup channels, and the initial priority sequence is obtained by sorting the scores in descending order.

[0028] Gigabit Ethernet is used as the primary channel, with multiple industrial wireless LANs and fieldbuses of different frequency bands deployed between the programmable logic controller (PLC) and industrial control computer (ICS) nodes. The `psutil` library in Python is used to access network interface status monitoring functions to obtain historical bit error rates (BERs). The `time` module records the occupancy time. A wireless signal time difference (TDD) positioning algorithm combined with Kalman filtering smoothing is used to calculate the distance between nodes. After obtaining these three parameters, the `MinMaxScaler` function from the Scikit-learn machine learning library is used to standardize the historical BER, occupancy time, and distance to a range of 0 to 1. A judgment matrix is ​​constructed based on the analytic hierarchy process (AHP) to calculate the normalized weights of each indicator, resulting in a weighted scoring formula for the backup channels. After the scoring is calculated, the built-in `sort` method in Python is called with the `reverse` parameter set to true to quickly sort the list of backup channel objects, thereby generating and storing an initial priority sequence dictionary in memory, ordered by weighted score in descending order.

[0029] In an optional embodiment, a weighted score is calculated based on the historical bit error rate, occupancy duration, and distance of the backup channel, including: The weighted score of the backup channel is calculated using the following formula: W = Ma × Eb × Tc × D; Where W is the dimensionless value of the weighted score, M is the preset dimensionless constant of the basic full score, E is the value of the channel historical bit error rate, a is the first preset conversion coefficient, T is the value of the total channel occupancy time in milliseconds, b is the second preset conversion coefficient, D is the value of the spatial three-dimensional straight-line distance of the node, and c is the third preset conversion coefficient.

[0030] During the initialization or periodic evaluation phase, the microcontroller extracts statistical parameters of the backup channel from the underlying communication log. The preset basic full-score dimensionless constant M is preferably set to 10000. The historical bit error rate E represents the ratio of erroneous data packets to total transmitted packets within the statistical period, with a preferred range of 0 to 500. The first preset conversion coefficient a is preferably set to 10.5. Simultaneously, the total channel occupancy time T is obtained by statistically calculating the cumulative number of milliseconds of actual data transmission and reception on the backup channel, with a reasonable value range of 0 ms to... Between ms, the corresponding second preset conversion coefficient b is preferably set to 0.0001; the spatial three-dimensional straight-line distance D of the node is obtained from the Euclidean distance in meters from the warehouse gateway to the unloading line terminal coordinates, with a preferred range of 5m to 500m, and the corresponding third preset conversion coefficient c is preferably set to 2.0. For example, for a backup channel with a statistically measured historical bit error rate E of 150, a total channel occupancy time T of 12,000,000 ms, and a spatial three-dimensional straight-line distance D of 80m, substituting the corresponding parameters into the weighting algorithm in the processor yields W=7065. The control system then uses this dimensionless benchmark score with high discrimination generated in real time by the computing unit to represent the channel quality. The backup channel weighted score has a linear negative correlation with the historical bit error rate, decreasing linearly as the bit error rate increases, such as Figure 2 As shown.

[0031] When a data interruption occurs in the primary channel, priority arbitration is triggered. The weighted score of each backup channel is multiplied by a correction factor that is positively correlated with the urgency of the task and negatively correlated with the offset of the corresponding cargo location of the backup channel. The backup channel with the highest correction score is selected for switching.

[0032] A daemon thread runs in the background of the industrial control computer, sending heartbeat packets with a period of 50ms via sockets. If no acknowledgment message is received for three consecutive cycles, or if the carrier loss flag is detected by calling the underlying netifaces library of the Linux system, the main channel data is determined to be interrupted, triggering a soft interrupt signal and activating the priority arbitration mechanism.

[0033] During the arbitration process, the application programming interface of the manufacturing execution system is invoked to read the delivery time difference of the current outbound task and calculate the task urgency level from an integer range of 1 to 10.

[0034] Simultaneously, the coordinates fed back from the machine vision positioning system or servo motor encoder are compared with the coordinates of the target standard cargo location to calculate the Euclidean distance and obtain the cargo location offset.

[0035] Numerical calculations were performed using the NumPy library to construct a correction factor. The numerator of this factor contains the task urgency multiplied by a constant proportional coefficient, and the denominator contains the cargo location offset plus a minimum value to prevent division by zero errors.

[0036] Traverse the initial priority sequence, multiply the initial weighted score of each backup channel by the corresponding correction factor, use the argmax function to find the network interface configuration parameters corresponding to the maximum value of the correction score, and modify the routing table by calling the operating system's ifconfig command and routing instructions to achieve hard switching of backup channels.

[0037] In an optional embodiment, the weighted score of each backup channel is multiplied by a correction factor that is positively correlated with the urgency of the task and negatively correlated with the offset of the corresponding cargo location of the backup channel, and the backup channel with the highest correction score is selected for switching, including: Extract the preset integer value representing the urgency level of the task from the main control task queue. Divide the preset integer value by the sum of the measured straight-line distance from the automated racking delivery point to the backup channel reference point and the preset zero-prevention constant to obtain a dimensionless quotient. Set the dimensionless quotient as the correction factor for the backup channel. Multiply the weighted score of the backup channel by the correction factor to obtain the correction score. Sort all backup channels in descending order of correction score and output the priority sequence. Select the backup channel with the highest priority for switching.

[0038] The system extracts preset integer values ​​representing the urgency level of tasks from the header of the master task queue messages issued by the manufacturing execution system. For example, it sets 10 for regular outbound tasks, 50 for urgent tasks, and 100 for express and urgent tasks. Simultaneously, it uses a laser rangefinder sensor to capture real-time data on the straight-line distance offset between the current outbound point of the automated storage and retrieval system and the target backup channel wireless access point reference point, e.g., a measured distance of 4.5m. The preset zero-prevention constant is preferably set to a fixed value of 0.5, and division logic is executed in the programmable logic controller to obtain a dimensionless quotient. Assuming the current emergency task parameter is set to 50, and the measured offset distance of the target backup channel is 4.5m, the calculated correction factor is 10. If the initial weighted score of the backup channel is 7065, the weighted score is multiplied by the correction factor to obtain a corrected score of 70650. The calculation is repeated for all links, and the corrected scores are stored in the form of a floating-point array. After being sorted in descending order by a quicksort algorithm, a pointer index array is generated. The control relay matrix immediately switches the network card transceiver interface hardware to the backup channel at position 0, achieving microsecond-level network takeover.

[0039] The verification module is used to use the XOR value of the last bit of the node's MAC address and the timestamp as a secondary arbitration token when an arbitration conflict occurs. The node with the smaller XOR value is given priority to access the network, and the Hamming weight of the XOR value is used to adjust the weight of the next round of occupancy time.

[0040] When the automated warehouse side node and the unloading line side node simultaneously send association requests to the same wireless access point, causing the backoff timer of the carrier sense multiple access conflict avoidance mechanism to reach full load, an arbitration conflict is triggered.

[0041] At this time, the microcontroller obtains the least significant byte of the node's media access control address by reading the network card register, and simultaneously calls the real-time clock module or network time protocol to obtain the least significant byte of the current system's microsecond-level timestamp.

[0042] The two are computed using the bitwise XOR operator to generate an 8-bit secondary arbitration token.

[0043] The token is embedded in the header of the media access control layer control frame and broadcast to the conflicting nodes. The microprocessor's arithmetic logic unit performs an unsigned integer comparison, and the node with the smaller token value immediately gains control of the radio frequency front end and establishes a communication link.

[0044] The Brian Kernighan algorithm is used to iteratively remove the least significant integer 1 from the binary representation of the XOR value, and the total number of iterations is counted to obtain the Hamming weight value.

[0045] The Hamming weight value is input into the preset exponential smoothing filter algorithm model. The weight adjustment step size is generated according to the linear mapping rule. In the next control cycle, the weight of the historical occupancy time index is subtracted from the step size, and all weights are normalized and redistributed to complete the adjustment of the scoring weights in the next round.

[0046] In an optional embodiment, the XOR value of the last bit of the node's MAC address and the timestamp is used as the secondary arbitration token, including: The last 8 bits of the MAC address of the hardware network card of the sending node are read using the underlying network protocol. The last 8 bits of the current millisecond-level clock timestamp variable of the sending device are extracted. After aligning the two, a bit-by-bit binary logical XOR operation is performed by the processor instruction. The calculated results of the 8 bits are concatenated to form a whole byte binary value, which is sent to the bus as a secondary arbitration token to perform arbitration allocation.

[0047] When the communication monitoring module detects a bus concurrency conflict between the automated storage and retrieval system (AS / RS) and the unloading line via carrier detection, the microcontroller's underlying driver immediately extracts the device's 48-bit unique MAC address from the network interface card (NIC) firmware register. It then uses a 0xFF mask to truncate the last byte of data. For example, if the last byte of the unloading line control gateway's MAC address is "5E" in hexadecimal, it is converted to binary and extracted as "01011110". Simultaneously, the bus synchronously accesses the kernel's real-time clock chip, retrieves the current millisecond-level Unix timestamp parameter, and performs a modulo-256 fast truncation operation to obtain the last 8 binary bits. Assuming the hexadecimal value of the last digit of this millisecond timestamp is "A3", its corresponding binary form is "10100011". The arithmetic logic unit is invoked to perform an XOR operation on two 8-bit binary numbers generated by hardware randomness and device characteristics, according to the rule that if they are the same, they are 0, and if they are different, they are 1. In the example above, the operation results in an integer binary sequence "11111101". This calculated integer is encapsulated in an unshielded micro-broadcast arbitration frame of fixed size 64 bytes and injected into the control bus. Each conflicting node extracts and compares the integer byte size in the received micro-frame and decides that the smaller value has priority, thereby achieving preemption separation and restarting the network session.

[0048] After the backup channel is switched, the transmitted message is segmented and hashed for verification. The segment length is directly proportional to the correction score and inversely proportional to the offset of the corresponding cargo location on the backup channel.

[0049] After the backup channel connection is established, the network protocol stack application layer takes over the data flow.

[0050] The default baseline message segment length is 512 bytes. The length is calculated by calling Python's math library. The length is equal to the baseline value multiplied by the corrected score, divided by the cargo location offset, and then rounded up to the nearest even number to meet the memory alignment requirements.

[0051] Based on the calculated segment length, the original application layer load is split using a slicing mechanism.

[0052] For each segment of data, the SHA256 secure hash algorithm object in the hashlib library is called to update the byte stream, and the hexdigest method is called to generate a 64-bit hexadecimal digest string, which is appended to the end of the corresponding segment's message as a check field.

[0053] After extracting the corresponding segment, the receiving end uses the same SHA256 algorithm to recalculate the hash value of the received payload, and compares it with the received check field using the built-in string comparison function. If they are completely consistent, the segment transmission is confirmed to be complete and error-free.

[0054] In an optional embodiment, the segment length is adjusted proportionally to the corrected score and inversely proportional to the offset of the corresponding cargo location on the backup channel, including: The basic message segmentation constant parameter is preset to 2048 bytes in the controller memory. The standard load byte length of a single segment packet in the channel transmission message is determined. The standard load byte length of a single segment packet is the maximum value between the preset positive integer minimum byte constant and the calculated value rounded down. The calculated value is the quotient of the dimensionless correction score of the backup channel divided by the distance compensation sum, multiplied by the basic message segmentation constant parameter. The distance compensation sum is the sum of the dimensionless value corresponding to the measured offset distance of the cargo position coordinate axis in meters for the backup channel and the preset dimensionless positive compensation constant.

[0055] Before processing large data packets, the data slicing module of the communication protocol stack locks memory, allocates a constant parameter of 2048 bytes for the basic packet segmentation, and declares the lower limit of the positive integer minimum byte constant used to avoid data fragmentation and network header overhead as 256 bytes.

[0056] For the currently determined backup communication path, the floating-point unit performs range normalization processing on the correction score of the backup channel, mapping it to the 0-2 interval to obtain a relative correction score coefficient, preferably 1.5. The actual absolute offset distance of the automated container in the guide rail direction is obtained through a coordinate encoder, for example, a measured value of 8.0m. This absolute offset distance is expressed as a dimensionless value and added to a preset dimensionless positive compensation constant, preferably 2.0, to obtain the distance compensation sum, for example, 8.0 + 2.0 = 10.0.

[0057] According to the system's preset calculation rules, the corrected scoring coefficient is divided by the distance compensation sum to calculate the quotient, for example, 1.5 / 10.0 = 0.15. This quotient is multiplied by the basic message segmentation constant parameter (2048 bytes) to obtain a floating-point intermediate value, for example, 307.2. The intermediate value is truncated using a floor function (such as Math.floor) to obtain a pure integer calculated value, i.e., 307. Further, the slicer compares this calculated value with the preset positive integer minimum byte constant safety threshold (256 bytes), and selects the maximum value of the two, i.e., 307 bytes, according to the highest validity principle, and sets it as the payload length quota for a single network segment packet. The network card driver then performs multi-packet reconstruction and distribution according to the quota template. Through this mechanism, the message segment length will dynamically decrease as the cargo location offset distance increases. When the offset distance exceeds the preset threshold, for example, 10m, the segment length stabilizes at the minimum byte constant. The variation pattern is described in [reference needed]. Figure 3 .

[0058] The control module is used to bind the outbound signal and the material feeding start / stop signal into a linkage message and divide it into sub-messages. The sub-messages are numbered from smallest to largest according to the data slice order. The receiving end requests retransmission based on the missing sub-message number and performs forward redundancy transmission on the preceding sub-messages. The number of redundancies increases as the initial sequence position increases.

[0059] The controller packages the high-level outbound pulse signal from the programmable logic controller and the start / stop Boolean signal of the unloading line motor into the same JSON-formatted linkage message string using the Protobuf serialization protocol.

[0060] Based on the previously determined segment length, the linkage message is divided into multiple sub-messages. An incrementing counter is used to assign an integer sequence number to each sub-message, starting from 0 or 1 and incrementing sequentially, as the sub-message header identifier.

[0061] The sending end invokes the transmission control protocol or the underlying transmission queue of a reliable transport protocol based on UDP to push the sub-message into the network.

[0062] The receiving end maintains a sliding window of expected sequence numbers. If the sequence number of an arriving sub-message is greater than the expected value, it determines that out-of-order delivery or packet loss has occurred and immediately sends a selective retransmission request control frame to the sending end, indicating the missing sub-message number.

[0063] Upon receiving a retransmission request, the sender not only retrieves the missing sub-message from the buffer pool and retransmits it, but also triggers a forward redundancy mechanism to send multiple copies of the previously confirmed sub-messages adjacent to the missing sub-message to resist sudden channel fading.

[0064] The sending end queries the index ranking of the backup channel in the initial priority sequence dictionary, and uses a lookup table method to determine the redundancy multiple based on the ranking value. That is, the later the ranking of the channel, the higher the probability of instability. The number of duplicate copies sent starts from one and increases linearly with the ranking. The retransmitted sub-message and the adjacent pre-acknowledged sub-message are packaged into multiple duplicate redundant frames and injected into the channel to resist sudden channel fading.

[0065] In an optional embodiment, the outbound signal and the material feeding start / stop signal are bundled into a linkage message and divided into sub-messages. The sub-messages are numbered in ascending order according to the data slice sequence, including: The multi-axis motion trajectory data stream of the three-dimensional container outbound action issued by the controller is extracted and written into the assembly buffer to represent the outbound signal. The continuous high-frequency sampled braking potential value sequence representing the motor and brake status of the downstream production line is extracted, filled into the buffer and appended after the data segment to complete the binding and encapsulation of the linkage message whose total data volume exceeds the single frame transmission limit. Extract the natural number 1 as the initial base number. When grouping the linked message into sub-messages, increment the natural number 1 step by step according to the data slice order to obtain a monotonically continuous integer and write it into the serial number slot of the first two bytes of the corresponding sub-message header as the number.

[0066] Furthermore, the control system acquires the three-dimensional motion trajectory position data stream of the automated container system during translation and lifting operations across the core via the memory access channel. This data stream is floating-point data acquired at a preset high-frequency interrupt interval of 5ms, forming a 1200-byte main outbound command stream and storing it in the assembly buffer. Simultaneously, the peripheral data acquisition gateway, within the same microsecond-level clock domain, acquires real-time drive voltage fluctuation data of the unloading line conveyor motor and millivolt-level braking potential data of the electromagnetic brake component through high-frequency sampling via analog-to-digital conversion circuit, forming a continuous data array of 2500 bytes. Within the assembly buffer, this data array is spliced ​​sequentially following the end of the trajectory data stream in a zero-fill mode, merging to form a giant linkage message with a total of 3700 bytes. Its size exceeds the Ethernet default maximum single-frame transmission unit limit of 1500 bytes, and is then sent to the message segmentation engine for fragmentation processing.

[0067] When the underlying software segments the aforementioned linked messages, a static auto-incrementing pointer variable, such as a uint16 type, is declared and initialized in the control register. Before encapsulating the first segmented sub-message, this pointer is assigned the preset base constant 1, thus writing the corresponding identifier (e.g., hexadecimal 0x0001) into the reserved segment of the first sub-message header. For subsequent data segments, the pointer variable is incremented using an adder hardware interrupt. Subsequent data segments also increment the pointer variable using adder hardware interrupts, assigning consecutively unique incrementing integer serial numbers to the second and subsequent sub-messages in the order of slicing.

[0068] In an optional embodiment, the preceding sub-messages are forward redundantly transmitted, with the number of redundancies increasing as the initial sequence position progresses, including: Retrieve the ranking index value of the currently used backup channel in the initial priority sequence, and calculate the total number of failed links of the backup channel that has failed in the current network link pool; The redundancy packet quota is calculated by adding the total number of faulty links to the ranking index value and then adding 1 to this sum. This quota instructs the network card to initiate the corresponding number of synchronous redundant transmissions of the preceding sub-packets.

[0069] Specifically, when the industrial communication monitoring agent component detects a message retransmission request or out-of-order void signal caused by electromagnetic interference, it will retrieve the static routing registry, extract the position of the currently active backup communication channel in the descending priority array, and convert it into an integer index value starting from 0, for example, an index value of 2. Simultaneously, the heartbeat process monitors the background keep-alive list, counts the number of offline faulty links that have experienced hardware interruptions and heartbeat timeouts, and thus determines the total number of faulty links in the current network environment. For example, if four wireless backup links are detected to be blocked, the total number of faulty links is four.

[0070] Based on the aforementioned fault degradation state, the core computing thread of the microcontroller adds the total number of faulty links (4) to the ranking index value (2) to obtain an intermediate parameter (6). A constant (1) is then added to this sum as a security protection threshold, resulting in a redundant packet transmission quota value of 7. The Ethernet Media Access Controller (MAC) is decoupled from the handshake restrictions of the upper-layer software protocol stack. It drives Direct Memory Access (DMA) and the network card's transmit register to perform hardware-level memory copying of missing sub-packets and adjacent pre-acknowledged sub-packets. Within microsecond-level idle time slots, it broadcasts a redundant copy of the original packet seven times consecutively without interval, according to the packet transmission quota value, reducing pulse interference and ensuring reliable delivery of the first segment of the signal.

[0071] If the number of retransmissions exceeds the set threshold, the weighted score will be reduced to the bottom. In the next cycle, the error rate weight will be temporarily increased and the scores will be re-sorted, and the linkage control will be executed.

[0072] At the sending end, a retransmission counter initialized to 0 is allocated for each sub-message, and the counter is incremented by 1 each time a retransmission request is executed.

[0073] The counter status is continuously monitored. When the logic judgment unit finds that the number of retransmissions of a certain sub-message is greater than the upper limit of five retransmissions in the preset configuration list, the core scheduler overwrites the weighted score of the current backup channel with a minimum value close to 0, moves it to the end of the priority dictionary list, and performs logical isolation.

[0074] The global timer is triggered, and the next millisecond-level scheduling cycle begins. The scheduler intervenes in the aforementioned hierarchical analysis judgment matrix. During the calculation process, the matrix eigenvector components of the historical bit error rate are multiplied by a gain factor of 2 to temporarily increase the weights. The newly normalized weight vector is then used to calculate the weighted score of all backup channels. The sort algorithm is called again to sort in descending order and establish a new backbone transmission link.

[0075] After the new link is established, the receiving end completes the lossless splicing of all linkage message slices, deserializes and parses the underlying action instructions, and drives the AC servo drive of the automated warehouse stacker and the frequency converter of the unloading line conveyor belt to start oscillating synchronously in a unified clock domain to complete the linkage control.

[0076] Ablation experiments were conducted in a smart warehouse with 20 nodes and 5 backup channels, deploying 6 microcontrollers with ranging sensors and 3 automated storage and retrieval systems (AS / RS). The communication network environment was set to a state of strong electromagnetic interference with a randomly injected network pulse error rate of 5%. Each experiment continuously ran 20,000 standard linked inbound and outbound tasks, focusing on comparing and testing the average task response latency, bus concurrency conflict resolution time, and data packet loss rate under extreme conditions. The experiment was divided into a baseline group, an ablation group, and a complete group. The baseline group adopted a traditional single fixed-priority channel evaluation mechanism and a fixed-length message slice transmission scheme. The ablation group selected channels by combining multi-factor weighted scoring with task correction factors, and added hardware-level MAC and timestamp XOR arbitration, but removed the segmented slice length adjustment and forward redundancy transmission mechanisms. The complete group fully adopted all the mechanisms proposed in this scheme, enabling linked message slice calculation and a forward redundancy cloning transmission mechanism based on ranking increments.

[0077] After seven days of continuous high-frequency operation, all experimental statistics were recorded and processed. Test results showed that the baseline group struggled to adapt to the interference in the workshop, with an average bus concurrent conflict resolution time of 245ms, an average task response latency as high as 510ms, and an overall link packet loss rate at a dangerously high level of 14.2%. The ablation group, benefiting from improved weighted scoring and underlying arbitration, rapidly reduced the average bus concurrent conflict resolution time to 38ms and the average task response latency to 182ms. However, due to the lack of redundant protection, the data packet loss rate remained at 5.4%. The complete group, after superimposing all protocol-based collaborative protection measures, exhibited the best reliability performance, not only reducing the concurrent conflict resolution time to 12ms and the average task response latency to 65ms, but also suppressing the overall link packet loss rate to 0.2% under severe electromagnetic interference. The key performance indicators of the three groups are compared as follows: Figure 4 As shown.

[0078] The above are all preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Therefore, all equivalent changes made in accordance with the structure, shape and principle of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A distributed signal transmission and linkage control system for automated warehouses and material unloading lines, characterized in that, include: The selection module is used to deploy a primary channel and multiple backup channels between the automated warehouse and the unloading line nodes. It calculates a weighted score based on the historical bit error rate, occupancy time, and the three-dimensional straight-line distance between the warehouse gateway and the unloading line terminal coordinates of the backup channels, and sorts them in descending order to obtain an initial priority sequence. When the primary channel experiences a data interruption, priority arbitration is triggered. The weighted score of each backup channel is multiplied by a correction factor that is positively correlated with task urgency and negatively correlated with the corresponding storage location offset. The backup channel with the highest correction score is selected for switching. The storage location offset is the straight-line distance between the automated warehouse's unloading point and the backup channel's reference point. The verification module is used to use the XOR value of the last bit of the node's MAC address and the timestamp as a secondary arbitration token when an arbitration conflict occurs. The network is accessed first based on the smaller XOR value, and the Hamming weight of the XOR value is used to adjust the weight of the occupancy time in the next round. After the backup channel is switched, the transmitted message is segmented and hashed for verification. The segment length is directly proportional to the correction score and inversely proportional to the offset of the corresponding storage location on the backup channel. The control module is used to bind the outbound signal and the material feeding start / stop signal into a linkage message and divide it into sub-messages. The sub-messages are numbered from smallest to largest according to the data slice order. The receiving end requests retransmission based on the missing sub-message number and performs forward redundant transmission of the preceding sub-messages. The number of redundant copies increases as the initial priority sequence position increases. If the number of retransmissions exceeds the set threshold, the weighted score is decayed to the last position. In the next cycle, the error rate weight is temporarily increased and the scores are re-sorted to execute linkage control.

2. The distributed signal transmission and linkage control system for the automated warehouse and unloading line according to claim 1, characterized in that, The weighted score is calculated based on the historical bit error rate, occupancy time, and three-dimensional straight-line distance between the warehouse gateway and the unloading line terminal. The weighted score of the backup channel is calculated using the following formula: W=Ma×Eb×Tc×D, where W is the dimensionless value of the weighted score, M is the preset dimensionless constant of the base full score, E is the value of the historical bit error rate of the channel, a is the first preset conversion coefficient, T is the value of the total occupancy time of the channel in milliseconds, b is the second preset conversion coefficient, D is the value of the three-dimensional straight-line distance between the warehouse gateway node and the unloading line terminal node, and c is the third preset conversion coefficient.

3. The distributed signal transmission and linkage control system for the automated warehouse and unloading line according to claim 1, characterized in that, The weighted score of each backup channel is multiplied by a correction factor that is positively correlated with the urgency of the task and negatively correlated with the offset of the corresponding storage location of the backup channel. The backup channel with the highest correction score is selected for switching. This process includes: extracting a preset integer value representing the urgency level of the task from the master task queue; dividing the preset integer value by the sum of the measured straight-line distance of the offset from the automated racking point to the backup channel reference point and a preset zero-prevention constant to obtain a dimensionless quotient; setting the dimensionless quotient as the correction factor for the backup channel; multiplying the weighted score of the backup channel by the correction factor to obtain the correction score; sorting all backup channels in descending order of correction score and outputting a priority sequence; and selecting the backup channel with the highest ranking for switching.

4. The distributed signal transmission and linkage control system for the automated warehouse and unloading line according to claim 1, characterized in that, The XOR value of the last bit of the node's MAC address and the timestamp is used as the secondary arbitration token. This process includes: reading the last 8 bits of the MAC address of the sending node's hardware network card using the underlying network protocol, extracting the last 8 bits of the current millisecond-level clock timestamp variable of the sending device, aligning the two bits, performing a bitwise binary logical XOR operation through processor instructions, concatenating the calculated results of the 8 bits to form a whole byte binary value, and sending it to the bus as the secondary arbitration token to perform arbitration allocation.

5. The distributed signal transmission and linkage control system for the automated warehouse and unloading line according to claim 1, characterized in that, The segment length is adjusted proportionally to the correction score and inversely proportional to the offset of the corresponding storage location of the backup channel. This includes: setting the basic message segmentation constant parameter to 2048 bytes in the controller memory; determining the fixed byte length of the single segment packet load of the channel transmission message; taking the maximum value between the preset positive integer minimum byte constant and the calculated value rounded down; where the calculated value is the quotient obtained by dividing the dimensionless correction score of the backup channel by the distance compensation sum and multiplying it by the basic message segmentation constant parameter; the distance compensation sum is the sum of the dimensionless value corresponding to the measured offset distance of the storage location coordinate axis in meters and the preset dimensionless positive compensation constant; and the offset distance of the storage location coordinate axis is the offset distance from the current shipping point of the automated rack to the reference point of the backup channel in the storage location coordinate axis direction.

6. The distributed signal transmission and linkage control system for the automated warehouse and unloading line according to claim 1, characterized in that, The outbound signal and the material unloading start / stop signal are bundled into a linkage message and divided into sub-messages. The sub-messages are numbered from smallest to largest according to the data slice order. This includes: extracting the multi-axis motion trajectory data stream of the three-dimensional container outbound action issued by the controller and writing it into the assembly buffer to represent the outbound signal; extracting the continuous high-frequency sampled braking potential value sequence representing the downstream production line motor and brake status, filling it into the buffer, and appending it after the data segment to complete the bundling and encapsulation of linkage messages whose total data volume exceeds the single frame transmission limit; extracting the natural number 1 as the initial base number, and gradually incrementing the natural number 1 according to the data slice order when grouping the linkage message into sub-messages to obtain a monotonically continuous integer and writing it into the serial number slot of the first two bytes of the corresponding sub-message header as the number.

7. The distributed signal transmission and linkage control system for the automated warehouse and unloading line according to claim 1, characterized in that, The preceding sub-messages are sent with forward redundancy. The number of redundancies increases as the initial priority sequence progresses. This includes: retrieving the ranking index value of the currently used backup channel in the initial priority sequence; calculating the total number of faulty backup channels in the current network link pool; adding the ranking index value to the total number of faulty links; and adding 1 to this sum to calculate the redundancy packet quota value, which instructs the network interface card to start synchronous redundancy transmission of the corresponding number of preceding sub-messages.

8. The distributed signal transmission and linkage control system for the automated warehouse and unloading line according to claim 1, characterized in that, Gigabit Ethernet is used as the primary channel, and multiple industrial wireless LANs and fieldbuses of different frequency bands are used as backup channels.

9. The distributed signal transmission and linkage control system for the automated warehouse and unloading line according to claim 1, characterized in that, When the automated warehouse side node and the unloading line side node simultaneously send association requests to the same wireless access point, causing the backoff timer of the carrier sense multiple access conflict avoidance mechanism to reach full load, an arbitration conflict is triggered.

10. The distributed signal transmission and linkage control system for the automated warehouse and unloading line according to claim 1, characterized in that, The Brian Kernighan algorithm is used to iteratively remove the least significant integer 1 from the binary representation of the XOR value, and the total number of iterations is counted to obtain the Hamming weight value.

Citation Information

Patent Citations

  • A CAN bus-based redundant system

    CN109639546B

  • Industrial Ethernet looped network protection system and method

    CN120915630A

  • Dynamic priority dual-mode communication fault-tolerant switching method and equipment

    CN121441822A