A kind of load cell intelligent guard system and method based on multi-node cooperation architecture

Abstract

The application relates to the technical field of automation control, and discloses a kind of earth weight intelligent guard system and method based on multi-node cooperation architecture, the system includes: initialization module;Measurement module;Backup module;Throttling module;Reconstruction module;Broadcast module.The method includes: allocating exclusive control token for bottom-layer peripheral equipment;Collecting on-site sub-node state vector and calculating performance degradation measure index;Generating summary to complete incremental data supplement transmission when below early-warning threshold;Triggering flow control to reduce business concurrency when reaching early-warning threshold;Revoking control authority to take over and issue subsequent driving instructions when reaching failure threshold;Sending messages between regional small total nodes to execute local offline closed loop when total node is offline.The application avoids peripheral equipment action conflict caused by concurrency, realizes distributed takeover and offline closed loop when node fails, and guarantees earth weight intelligent guard business continuity and data security based on multi-node cooperation architecture.

Description

Technical Field

[0001] This invention relates to the field of automation control technology, specifically to a weighbridge intelligent monitoring system and method based on a multi-node collaborative architecture. Background Technology

[0002] Weighbridges are core large-scale weighing instruments used in the logistics and mining industries for weighing large quantities of materials from vehicles. As companies increasingly demand automated management, the traditional method of manual registration and document processing is gradually being replaced due to its inefficiency. Introducing intelligent systems that integrate hardware and software to monitor the weighing site has gradually enabled unmanned management from scheduled entry to weighing exit.

[0003] For unmanned applications, existing technologies already have publicly available solutions. For example, Chinese patent application CN121655664A proposes an unattended intelligent weighbridge weighing method and device, which integrates access control, weighing, and video capture through a unified entry architecture to prevent cheating behaviors such as falsely reporting tare weight or gross weight. Another example is Chinese patent application CN120141629A, which corrects measurement errors caused by vehicle position deviations by setting positioning points and monitoring points. These existing technologies effectively improve the accuracy and process standardization of single weighing by introducing automated hardware.

[0004] Current conventional automation applications typically deploy multiple terminal computers distributed across the site, connecting underlying devices such as collision avoidance radar, license plate capture cameras, and barrier gates to a central server via a network. Business logic is centrally processed by the central node, which then issues control commands to achieve centralized data storage and coordinated operation of the on-site hardware. Some existing technologies (such as Chinese patent application CN121412860A) attempt to introduce edge computing architectures for localized data verification to mitigate cloud latency, but their overall data flow and control logic still heavily rely on a top-down vertical network hierarchy.

[0005] However, this application model has limitations under complex operating conditions. In networks with multiple terminal devices communicating collaboratively, the lack of exclusive control and dynamic load measurement mechanisms for the underlying physical links makes concurrent requests prone to peripheral device conflicts and system overload crashes. Furthermore, this centralized architecture heavily relies on the backbone network. When faced with partial network outages or core node failures, it lacks distributed breakpoint takeover, incremental backup, and offline verification closed-loop strategies, leading to paralysis of the field control link and loss of metering data, making it difficult to guarantee continuous business operation under extreme conditions.

[0006] Therefore, this invention proposes a weighbridge intelligent monitoring system and method based on a multi-node collaborative architecture to address the shortcomings of existing technologies. Summary of the Invention

[0007] To address the shortcomings of existing technologies, this invention provides a weighbridge intelligent monitoring system and method based on a multi-node collaborative architecture. This solves the problems of existing centralized weighbridge systems, which are prone to underlying device control conflicts under concurrent requests from multiple terminals, and lack of distributed takeover and offline closed-loop mechanisms that lead to paralysis of on-site operations when there are network anomalies or core node failures.

[0008] To address the above problems, the present invention provides the following technical solution:

[0009] The first aspect of this invention provides a weighbridge intelligent monitoring system based on a multi-node collaborative architecture, comprising:

[0010] The initialization module is used to distribute global configuration parameters to regional sub-nodes and field sub-nodes to establish the calculation benchmark for weighbridge measurement, and to allocate exclusive control tokens to the underlying peripherals of the weighbridge to establish access restrictions; the global configuration parameters include a warning threshold and a failure threshold, and the failure threshold is greater than the warning threshold;

[0011] The measurement module is used to collect the multi-dimensional operating status vector of the field sub-nodes and calculate the performance degradation measurement index of the field sub-nodes based on the multi-dimensional operating status vectors.

[0012] The backup module is used to perform routine weighbridge measurement to generate weighing data according to the global configuration parameters when the performance degradation metric index is lower than the warning threshold, generate a data integrity verification summary for the weighing data, and use the data integrity verification summary to perform incremental data retransmission.

[0013] The throttling module is used to trigger a flow limiting control signal when the performance degradation metric index reaches the warning threshold and is lower than the failure threshold.

[0014] The refactoring module is used to revoke the control authority of the field sub-node and transfer the exclusive control token to the regional small master node to take over and issue subsequent weighbridge business driving instructions when the performance degradation metric index reaches the failure threshold.

[0015] The broadcast module is used to send a message carrying the hash value of the digital certificate between the regional sub-nodes if the central node is offline, and to perform a local offline weighing closed loop based on the received message.

[0016] Furthermore, the initialization module is specifically used for:

[0017] An authentication signature value is generated based on the media access control address of the network interface device, the network interconnection protocol address of the field sub-node, and the current timestamp of the central node.

[0018] The authentication signature value and the current timestamp are combined to form the exclusive control token and sent to the field sub-node;

[0019] The authentication signature value, the current timestamp, and the network interconnection protocol address are written into the access control list register inside the network interface device for comparison and restriction, so as to allocate the exclusive control token to the underlying peripheral device.

[0020] Furthermore, the multidimensional runtime state vector includes network request response latency, memory usage, and the length of the business instruction pending queue;

[0021] The measurement module is specifically used for:

[0022] The actual collected network request response latency is combined with the normal lower limit and the extreme upper limit value to perform extreme value standardization and boundary truncation calculations;

[0023] The results of the extreme value standardization operation and the boundary truncation operation are weighted and summed with the normalized memory usage and the length of the business instruction pending queue, respectively, and the output scalar result is used as the performance degradation metric index.

[0024] Furthermore, when scheduling processing logic based on the range of the performance degradation index, the weighbridge intelligent monitoring system incorporates a state switching hysteresis mechanism: when the performance degradation index rises from the healthy range and crosses the warning threshold for the first time, the system immediately triggers the throttling module to execute flow restriction; when the node load eases and the performance degradation index falls back, it is necessary to determine that the index is below the warning threshold minus the preset hysteresis bias for several consecutive sampling periods before the flow restriction control can be lifted and normal business can be restored, thereby avoiding frequent fluctuations in business status caused by small fluctuations in node performance at the threshold edge.

[0025] Furthermore, the backup module is specifically used for:

[0026] The weighing data is recorded as a write-ahead log on a local storage device;

[0027] Extract the binary byte sequence after serializing the weighing data;

[0028] The basic load is constructed based on the auto-incrementing sequence number in the write-ahead log and the random salt constant configured inside the field sub-nodes.

[0029] A fixed-length string is generated from the underlying payload using a secure hash algorithm to serve as a data integrity verification digest.

[0030] Furthermore, the backup module is specifically used for:

[0031] Periodically read the current record cursor of the write-ahead log;

[0032] Compare the current record cursor with the latest received cursor issued by the regional sub-node;

[0033] The set of unsynchronized data that falls within the difference range between the current record cursor and the latest received cursor is selected for the incremental data retransmission operation.

[0034] Furthermore, the throttling module is specifically used for:

[0035] Send a flow control signal carrying a pause release flag to the source gate location to which the on-site sub-node belongs, so as to suspend the registration of new vehicles;

[0036] Modify the dispatching parameters of the manufacturing execution system to delay the allocation of the arrival time slices of reserved vehicles using the calculated time offset;

[0037] A control frame containing hardware timing parameters is generated and sent to a logic programmable gate array to increase the gate opening delay interval.

[0038] Furthermore, when calculating the time offset, the throttling module:

[0039] Calculate the difference between the current performance degradation metric index and the warning threshold, and divide it by the difference between the failure threshold and the warning threshold to obtain a quotient.

[0040] The quotient is multiplied by a smoothing coefficient and then used as the exponential part of the natural constant.

[0041] The time offset is obtained by multiplying the exponential portion by the reference delay constant.

[0042] Furthermore, the reconstruction module is specifically used for:

[0043] A write command carrying a management signature is issued to clear the original address bound in the access control list register inside the network interface device, thereby revoking the control permissions of the field sub-node.

[0044] The exclusive control token is transferred to the regional sub-monitor node to achieve the takeover;

[0045] Send status query messages to the underlying peripheral device to collect the level and digital signals of the device operating parameters;

[0046] The level and digital signals are normalized to construct a field environment state vector. An Euclidean distance algorithm is then used to map and compare this field environment state vector to a business logic breakpoint to determine the location for issuing subsequent drive commands. Preferably, when using the Euclidean distance algorithm for comparison, the distance between the current field environment state vector and the baseline state vectors of each standard business stage is calculated by summing the square differences and then taking the square root. The business stage containing the baseline state with the smallest calculated result is selected as the determined business logic breakpoint to continue executing the field peripheral drive actions.

[0047] Furthermore, when the reconstruction module uses the Euclidean distance algorithm to determine the breakpoint of the business logic, it introduces an abnormal state fallback strategy: after calculating the distance values ​​between the current field environment state vector and all reference state vectors, if it is determined that all distance values ​​are greater than the preset fault tolerance distance limit, or if there are multiple different reference states with the same distance value and all of them are the minimum value, resulting in a matching conflict, then it is determined that the underlying peripheral device has physical damage or unknown abnormality. At this time, the system actively interrupts the automatic takeover process and intercepts the issuance of subsequent drive commands, and simultaneously triggers audible and visual alarms and data freezing to prevent irreversible erroneous metering accounts.

[0048] Furthermore, the broadcast module is specifically used for:

[0049] The data packet containing the hash value of the digital credential is sent to other regional mini-nodes via LAN multicast using the User Datagram Protocol.

[0050] The regional sub-node that receives the data packet searches the local receiving buffer pool for the corresponding entry based on the current physical identity identifier collected by the weighbridge's underlying peripheral device.

[0051] The entry timestamps in the corresponding entries are extracted, recalculated, and compared. When a match is found, the control command for the offline weighing closed loop is automatically issued.

[0052] Furthermore, when calculating the hash value of the digital credential, the broadcast module concatenates the physical identity string of the target vehicle, the entry timestamp value, and the peer authentication key constant pre-embedded within the regional mini-node. Then, it inputs the string into a secure hash algorithm calculation function to generate a digest value, so as to maintain the tamper-proof transfer and weighing closed loop of vehicle identity in the local area under the condition of backbone network blockage.

[0053] Furthermore, when the broadcast module sends data packets containing digital credential hash values ​​via LAN multicast, it appends a unique identifier sequence number and a lifetime hop count feature to the header of the data packet. Before pushing the data packets into their local receive buffers, other regional nodes that receive the data packets first extract the unique identifier sequence number for interception and deduplication, and decrement the lifetime hop count during message forwarding. This effectively suppresses broadcast storms within the underlying LAN during backbone network outages, ensuring the availability of internal communication bandwidth under offline closed-loop conditions.

[0054] A second aspect of this invention provides a method for intelligent monitoring of weighbridges based on a multi-node collaborative architecture, applied to the intelligent monitoring system of weighbridges described in the first aspect. The method includes the following steps:

[0055] The initialization module distributes global configuration parameters during the system startup phase and calculates and issues control credentials to allocate exclusive control tokens to underlying peripherals.

[0056] The measurement module obtains a multi-dimensional operating status vector composed of device network and hardware parameters during normal operation, and calculates the performance degradation measurement index of field sub-nodes.

[0057] When the performance degradation metric index is lower than the warning threshold, the backup module performs routine business, generates a data integrity verification digest using a hash function, and completes incremental data retransmission.

[0058] When the performance degradation metric reaches the warning threshold, the throttling module intervenes in the upstream software platform's dispatch scheduling and controls the physical execution mechanism to increase latency, triggering rate limiting control to reduce business concurrency.

[0059] When the performance degradation metric index reaches the failure threshold, the refactoring module takes over the authority, obtains and uses the spatial distance to compare the on-site environmental state vector and then issues subsequent driving instructions.

[0060] When the central node is offline, the local area network autonomous process is initiated through the broadcast module, and the digital certificate hash value is horizontally multicast among peer nodes to execute the local offline closed loop.

[0061] This invention provides a weighbridge intelligent monitoring system and method based on a multi-node collaborative architecture, which has the following beneficial effects:

[0062] This invention, based on hardware exclusive access restrictions and real-time performance degradation measurement, achieves hierarchical transfer of control. When healthy, it ensures data synchronization and tamper-proofing; under high load, it proactively limits and reduces voltage; when a node fails, it forcibly transfers control to a higher level; and when the backbone network is down, it completes offline closed-loop operation via local area network communication. This collaborative architecture eliminates the risks of concurrency conflicts and single-point failures common in traditional centralized control systems. It requires no manual intervention under abnormal hardware and software conditions, ensuring continuous operation and data consistency of weighbridge operations.

[0063] This invention assigns an exclusive control token containing a signature and timestamp to the network interface device through an initialization module, and writes it into an internal register for comparison when receiving data frames. This mechanism establishes exclusive device access restrictions in a multi-node network environment, intercepting abnormal concurrent requests and avoiding action conflicts caused by multiple commands. This ensures the secure response of the weighbridge measuring hardware and provides the underlying control foundation for realizing intelligent device monitoring under a collaborative architecture.

[0064] This invention records weighing data as a write-ahead log through a backup module, uses the difference between the local and upstream cursors to filter out out-of-synchronization data and perform incremental retransmission, and combines a hash algorithm to generate a tamper-proof digest. This backup method prevents the risk of tampering during transmission and ensures closed-loop synchronization of weighbridge historical data after the backbone network is restored. This approach meets the needs of independent data storage for multiple nodes and improves the security of intelligent monitoring of business data under a collaborative architecture.

[0065] This invention monitors node status and calculates a degradation metric index through a measurement module. When the index reaches a warning threshold, it triggers throttling control. By delaying upstream order dispatch time slices, sending suspension registration signals, and increasing the delay of gate operation, it proactively reduces the amount of concurrent requests flowing into the site. This throttling strategy reduces the instantaneous load caused by multi-node concurrency, avoids resource exhaustion in the collaborative architecture during peak business periods, and maintains the continuous operation of the intelligent weighbridge monitoring system.

[0066] This invention utilizes a reconfiguration module to transfer control tokens when a node fails and calculates the on-site environmental state vector to compare with business logic breakpoints to continue issuing commands. A broadcast module multicasts digital credentials to target nodes for offline control when the main node is offline. This hierarchical switching mechanism maintains the status takeover and regional closed loop of the weighbridge's on-site peripherals even under network outage conditions. This disaster recovery configuration meets the independent takeover requirements of multiple nodes, ensuring the normal operation of intelligent monitoring services in offline mode under the collaborative architecture. Attached Figure Description

[0067] Figure 1 This is an architecture diagram of a weighbridge intelligent monitoring system based on a multi-node collaborative architecture, according to an embodiment of the present invention.

[0068] Figure 2This is a flowchart of a weighbridge intelligent monitoring method based on a multi-node collaborative architecture according to an embodiment of the present invention.

[0069] Figure 3 This is a flowchart illustrating the system startup phase and exclusive control token allocation process according to an embodiment of the present invention.

[0070] Figure 4 This is a flowchart of the operation status acquisition and performance degradation measurement calculation according to an embodiment of the present invention;

[0071] Figure 5 This is a flowchart illustrating the routine measurement and data backup and retransmission process for health intervals according to an embodiment of the present invention.

[0072] Figure 6 This is a flowchart of the current limiting control intervention under high load conditions according to an embodiment of the present invention;

[0073] Figure 7 This is a flowchart illustrating the process of permission bypassing and service recovery in the event of node failure, as described in an embodiment of the present invention.

[0074] Figure 8 This is a flowchart illustrating the local area network autonomy and offline closed-loop operation under backbone network blocking, as described in this embodiment of the invention.

[0075] Figure 9 This is a graph showing the performance degradation metric index and the dynamic response curve of the delay time slice in an embodiment of the present invention.

[0076] Figure 10 This is a line graph showing the 24 / 7 operation status of the intelligent weighbridge monitoring system nodes according to an embodiment of the present invention.

[0077] Among them, 101 is the initialization module; 102 is the measurement module; 103 is the throttling module; 104 is the reconstruction module; 105 is the broadcast module; and 106 is the backup module. Detailed Implementation

[0078] The technical solutions in 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.

[0079] Please see Figure 1 This invention provides a weighbridge intelligent monitoring system based on a multi-node collaborative architecture, which may include: an initialization module 101, a measurement module 102, a throttling module 103, a reconstruction module 104, a broadcast module 105, and a backup module 106.

[0080] The intelligent weighbridge monitoring system adopts a three-layer network and logical topology deployment: the top layer deploys the central node, the middle layer deploys the regional nodes, and the bottom layer deploys the field nodes.

[0081] The field nodes directly interact with the underlying peripherals, which include collision avoidance radar, weighbridge sensors, and barrier gates. Data communication links between the central node, regional sub-nodes, and field nodes are established via industrial Ethernet and message queue buses. The underlying peripherals connect to the local area network (LAN) through network interface devices. In one embodiment, the network interface device is an industrial Ethernet input / output module or a serial server.

[0082] During the startup phase of the weighbridge intelligent monitoring system, the initialization module 101 configured at the central node generates global configuration parameters and distributes them to regional sub-nodes and field sub-nodes. At the same time, the initialization module 101 assigns exclusive control tokens to the underlying peripherals to establish access restrictions for network interface devices.

[0083] During the normal operation phase of the weighbridge intelligent monitoring system, the measurement module 102 configured at the regional sub-node collects multi-dimensional operating status vectors of the field sub-nodes. The measurement module 102 calculates the performance degradation measurement index of the field sub-nodes based on the multi-dimensional operating status vectors. The weighbridge intelligent monitoring system schedules the corresponding processing logic to execute the corresponding process according to the interval range where the performance degradation measurement index is located.

[0084] When the performance degradation metric index falls below the warning threshold, the field sub-node performs routine material metering operations. The backup module 106 configured at the field sub-node records the structured weighing data as a write-ahead log on the local storage device. The backup module 106 uses a hash function to generate a data integrity verification digest and asynchronously transmits incremental data to the regional sub-node.

[0085] When the performance degradation metric index rises and reaches the warning threshold, the throttling module 103 triggers flow restriction control. The throttling module 103 sends a flow restriction control signal to the source gate of the field sub-node. The throttling module 103 modifies the dispatching parameters of the manufacturing execution system to delay the allocation of the entry time slice of the reserved vehicle. In addition, the throttling module 103 issues an instruction to increase the gate opening delay interval, thereby reducing the business concurrency flowing into the field sub-node.

[0086] When the performance degradation metric index rises further and reaches the failure threshold, the refactoring module 104 takes over the authority, forcibly revokes the control authority of the field sub-node, and transfers the exclusive control token to the regional small master node.

[0087] After obtaining the exclusive control token, the reconfiguration module 104 sends a status query message to the network interface device. The reconfiguration module 104 collects the current status of the anti-collision radar, the weighbridge sensor and the barrier gate to form a field environment status vector. The reconfiguration module 104 maps and compares the field environment status vector to the business logic breakpoint. Then, the reconfiguration module 104 takes over the field sub-node and sends subsequent drive instructions according to the determined business logic breakpoint.

[0088] If the central node goes offline due to network communication interruption, the broadcast module 105 initiates the local area network autonomous process. The broadcast module 105 sends messages between regional sub-nodes using local communication channels. The broadcast module 105 horizontally delivers the hash value of the target vehicle's digital certificate to the regional sub-nodes in the target area. The regional sub-nodes that receive the hash value of the digital certificate compare it with the local cached features and execute a local offline closed loop to complete the weighing operation.

[0089] In other alternative embodiments, the underlying peripherals also include an intelligent video analytics camera and an infrared grating array. The intelligent video analytics camera uses a built-in target detection algorithm (e.g., YOLOv8) to extract the coordinate information of vehicles and personnel, and combines it with a temporal action recognition model to output behavioral feature signals (including vehicle not fully on the weighbridge and personnel abnormally getting off the vehicle). Simultaneously, the infrared grating arrays distributed on both sides of the weighbridge control the through-beam optical paths of the transmitter and receiver via a microcontroller, collecting the occlusion and brightness states of each channel, thereby outputting vehicle position and contour feature signals. These behavioral feature signals and contour feature signals are integrated as additional Boolean control quantities into the field environment state vector of the reconstruction module 104. When the system triggers the authorization takeover process, if the reconstruction module 104 detects that the Boolean control quantity is valid, it interrupts the distance comparison calculation of the state vector and issues a control command to freeze the field hardware actions. Furthermore, the field sub-nodes can generate a behavioral score based on the above feature signals and use this score as a basic payload in the string concatenation calculation of the digital credential hash value in the broadcast module 105 to maintain the synchronous verification of vehicle identity and compliance status under local area network autonomous operation.

[0090] In other alternative embodiments, the underlying peripherals also include a solid-state lidar and a micro-weather station. The solid-state lidar is used to collect on-site 3D point cloud data and extract vehicle attitude and centering coordinates; the micro-weather station is used to collect on-site temperature, humidity, wind speed, and wind pressure values ​​in real time. When the on-site sub-node detects that the signal-to-noise ratio of the feedback signal from the collision avoidance radar is lower than the calibrated threshold, the 3D point cloud data of the solid-state lidar is automatically switched as an alternative input source. When determining the business logic breakpoint, the reconstruction module 104 extracts the geometric feature components of the above-mentioned 3D point cloud data to expand the dimension parameters of the on-site environmental state vector and participates in the Euclidean distance calculation process. At the same time, the system extracts the temperature value output by the micro-weather station and introduces it into the tare calculation stage of the weighbridge sensor to perform real-time zero-point drift compensation; the throttling module 103 extracts the wind speed value output by the micro-weather station and converts it into a wind resistance compensation coefficient, which is superimposed on the hardware timing parameters of the underlying control frame to extend the gate opening delay interval under high wind load conditions.

[0091] Please see Figure 2 This invention provides a method for intelligent monitoring of weighbridges based on a multi-node collaborative architecture, comprising the following steps:

[0092] S1 distributes global configuration parameters and allocates exclusive control tokens to underlying peripherals through initialization module 101;

[0093] S2, calculates the performance degradation metric index of the field sub-nodes through the metric module 102;

[0094] S3, when the performance degradation metric index is lower than the warning threshold, the backup module 106 performs regular business and completes incremental data retransmission.

[0095] S4, When the performance degradation metric index reaches the warning threshold, the throttling module 103 triggers rate limiting control to reduce the business concurrency;

[0096] S5, when the performance degradation metric index reaches the failure threshold, the refactoring module 104 performs permission takeover and issues subsequent driving instructions.

[0097] S6, when the central node is offline, initiates the local area network autonomous process through the broadcast module 105 to execute the local offline closed loop.

[0098] The technical solution of the present invention will be further described below with reference to the accompanying drawings.

[0099] See Figure 1 and Figure 3During the startup phase of the intelligent weighbridge monitoring system, it is necessary to ensure that nodes at all levels have a unified business benchmark and to prevent underlying devices from being falsely triggered by multiple concurrent requests. To this end, the initialization module 101 distributes global configuration parameters and assigns exclusive control tokens to underlying peripherals, thereby establishing access restrictions for network interface devices. The specific execution details in this embodiment include the following sub-steps:

[0100] S11, the initialization module 101 extracts business rule elements from the relational database management system to generate global configuration parameters.

[0101] The initialization module 101 accesses a locally deployed relational database management system. The relational database management system is a well-known data storage software platform used for persistent storage and centralized management of basic data related to weighbridge business rules. As a preferred approach, the initialization module 101 uses Structured Query Language (SCL) commands to read preset material billing standard rate tables, license plate recognition error-tolerant matching rules, weighbridge sensor calibration error range values, and node status evaluation thresholds. These preset tables, rules, error range values, and evaluation thresholds are all baseline data manually entered into the relational database management system in advance by business management personnel based on actual operational needs. SCL commands are standard code statements that directly interact with the database for data extraction. For example, the initialization module 101 accurately locates and extracts the corresponding material billing standard rate table by executing the specific query statement `SELECT * FROM fee_rule`.

[0102] Specifically, the material billing standard rate table specifies, for example, a billing benchmark parameter of 50 yuan per ton of steel; the license plate recognition fault-tolerant matching rules specify, for example, a fault-tolerant comparison threshold that allows for single-character optical character recognition errors; the weighbridge sensor calibration error range specifies, for example, a weight drift of ±5 kg allowed under no-load conditions; and the node status evaluation threshold specifies the threshold used to determine the load status of on-site sub-nodes and trigger corresponding flow limiting control or authorization takeover.

[0103] To facilitate efficient parsing across different underlying platforms in industrial Ethernet, the initialization module 101 formats and converts the read multi-type business rule elements to construct a set of key-value pairs with a unified data structure. The integrated set of key-value pairs is the global configuration parameter.

[0104] For example, the system can use a lightweight JSON format to encapsulate these elements, containing specific key-value pairs representing the standard rate and attribute fields representing the number of tolerance bits in the generated JSON text. In this way, the globally configured parameters explicitly define the unified calculation benchmark for material measurement, billing, and data verification performed by the intelligent weighbridge monitoring system during subsequent routine operation.

[0105] S12, the initialization module 101 uses the publish-subscribe mechanism of the message queue bus to distribute global configuration parameters to the regional sub-nodes and field sub-nodes, and determines whether a response has been received from a specific node. If so, a point-to-point data retransmission operation is triggered. If the result is negative, it indicates that a response frame has been received and proceeds to step S13.

[0106] After generating the global configuration parameters, the initialization module 101 connects to the message queue bus and publishes a data packet containing the global configuration parameters to a predefined topic communication channel (a predefined topic communication channel refers to a data channel identifier specifically configured in the message middleware for parameter distribution, for example, setting the channel identifier name to global_config_topic). Regional sub-nodes and field sub-nodes listen to this topic communication channel as subscribers during the startup phase. Upon receiving the data packet, they parse the packet payload and extract the global configuration parameters, which are then stored locally. Due to transient network packet loss caused by electromagnetic interference in the industrial field, the subscriber will return an acknowledgment frame via the message queue bus after storing the parameters. If the initialization module 101 does not receive a response from a specific node within a set timeout window (e.g., a pre-set response waiting time of 500 milliseconds), it triggers a point-to-point data retransmission operation for that node to avoid a logical dead zone caused by missing configuration data leading to the stagnation of metering services. For the data encapsulation and connection establishment mechanism of the underlying message queue protocol, those skilled in the art can use existing transmission control protocol components; the connection handshake process is a well-known technology in the field and will not be described further here.

[0107] S13: An exclusive control token is calculated and generated by the initialization module 101, and then issued or written to the access control list register inside the network interface device to establish access restrictions. The network interface device determines whether the exclusive control token has expired due to the passage of time. If the determination result is yes, the field node re-initiates an application to the central node to obtain an updated exclusive control token; if the determination result is no, the token has not expired, and the pass-through drive level signal is allowed to enforce the access restrictions of the network interface device.

[0108] Specifically, to prevent multiple field nodes from sending action commands to the same underlying peripheral device concurrently, which could lead to action conflicts or even device damage, the initialization module 101 sets permission isolation settings for the network interface devices connected to the anti-collision radar, weighbridge sensor, and barrier gate.

[0109] The initialization module 101 reads the media access control address of the network interface device and combines it with the network interconnection protocol address of the field node currently designated to take over the underlying peripheral and the current timestamp of the central node to form a basic authentication payload. The initialization module 101 performs an encrypted hash operation on the basic authentication payload to generate a fixed-length authentication signature value. This authentication signature value and the current timestamp of the central node together constitute an exclusive control token. The exclusive control token logically represents a digital credential that a specific field node has unique and legitimate driving rights to a specific network interface device.

[0110] Specifically, the initialization module 101 calculates the authentication signature value of the exclusive control token. The following formula is used for calculation:

[0111] ;

[0112] In the formula, This is a hash-based message verification code calculation function. A 256-bit length is chosen for the secure random string key configured within the central node to ensure anti-counterfeiting security. The media access control address set at the factory for network interface devices; The network interconnection protocol address bound to the target site sub-node; The current timestamp of the central master node is set to millisecond precision; This is a string concatenation operator.

[0113] The initialization module 101 sends the generated exclusive control token to the corresponding field sub-node for storage, and simultaneously sends the authentication signature value contained in the exclusive control token. The current timestamp of the central node and the network interconnection protocol address of the on-site sub-node. Write it into the access control list register inside the network interface device.

[0114] When a network interface device receives an external network data frame, it triggers its internal rule matching logic. The network interface device extracts the source address and accompanying signature payload from the network data frame header and compares them at the hardware level with preset values ​​in the internal access control list register. If the source address and signature of the network data frame perfectly match the record characteristics of the exclusive control token, and the current comparison time does not exceed the current timestamp of the central master node recorded in the register, the matching is successful. Only when the limit defined by the preset validity period (such as a two-hour fixed duration limit set in advance by security personnel) is superimposed, will the network interface device transmit the drive level signal to the anti-collision radar, weighbridge sensor or barrier gate.

[0115] If the token expires over time, the control request from the field node will be rejected. In this case, the field node needs to re-submit a request to the central node to obtain a new exclusive control token and maintain the normal control link. Through the above-mentioned interception operation based on hardware register comparison, the initialization module 101 assigns exclusive control permissions to the underlying peripherals and establishes access restrictions for network interface devices at the physical link layer.

[0116] See Figure 1 and Figure 4 During the normal operation phase of the weighbridge intelligent monitoring system, the measurement module 102 collects various indicators reflecting the underlying status of the nodes, thereby calculating the performance degradation measurement index of the on-site sub-nodes. Based on the numerical range of this index, the weighbridge intelligent monitoring system will schedule the corresponding processing logic to execute the corresponding process. The specific execution details are divided into the following sub-steps:

[0117] S21, the measurement module 102 collects various indicator parameters reflecting the underlying operating status of the node to form or combine them into a multi-dimensional operating status vector. These indicator parameters include: network request response latency, memory usage, and business instruction pending queue length.

[0118] Specifically, in order to monitor the real-time communication and load status of each node, the measurement module 102 configured at the regional sub-node initiates periodic polling of the field sub-nodes via industrial Ethernet. The measurement module 102 sends Internet Control Message Protocol (ICP) probe packets to the field sub-nodes via industrial Ethernet and records the time difference between sending the probe packet and receiving the corresponding reply packet, thereby obtaining the network request-response latency.

[0119] Because of occasional network disconnections in industrial sites, if the measurement module 102 does not receive a response message within a preset timeout window (e.g., two thousand milliseconds), the network request response delay will be directly assigned a preset maximum penalty value (e.g., nine thousand nine hundred and ninety-nine milliseconds). This maximum penalty value is a set limit delay constant, specifically used to replace missing data when a network disconnection prevents the acquisition of the true delay, so as to ensure that the normalized result of the network delay term is forced to output the maximum deterioration state with full bias in subsequent calculations, thus avoiding the acquisition process from entering a deadlock state of infinite waiting.

[0120] Meanwhile, the measurement module 102 uses the Simple Network Management Protocol to read the resource registers of the field node operating system (here, the operating system refers to the basic software environment that supports the operation of the field node business programs, such as an industrial-grade Linux platform) to obtain the number of bytes of physical memory currently occupied, which is the memory usage.

[0121] Furthermore, the measurement module 102 queries the thread pool status of the internal business programs of the field sub-nodes to count the number of business requests that have not yet been completed and are in a queue, thereby obtaining the length of the business instruction suspension queue. For the specific communication handshake process of the aforementioned Internet Control Message Protocol and Simple Network Management Protocol, those skilled in the art can refer to relevant network communication standards and specifications; the implementation of their protocol stack code is well-known in the field and will not be elaborated upon here.

[0122] After collecting the three basic status data points mentioned above, the measurement module 102 combines the acquired network request response latency, memory usage, and business instruction pending queue length into a one-dimensional data matrix in a fixed order. This one-dimensional data matrix is ​​the multi-dimensional operational status vector. By extracting data from these three independent dimensions, the multi-dimensional operational status vector records the real-time operational status of the field nodes from three levels: network connectivity, hardware resource consumption, and software concurrent load.

[0123] S22, the performance degradation metric index of the field sub-node is calculated by the metric module 102 based on the acquired multi-dimensional operating state vector.

[0124] Since the dimensions and orders of magnitude of the various index parameters contained in the multidimensional operating state vector are different, directly accumulating the values ​​would mask the changes in smaller numerical indicators. The measurement module 102 needs to use an extreme value normalization method to map each indicator to a dimensionless interval of zero to one. Based on this, a weighted summation is performed by assigning appropriate weight coefficients to each normalized indicator. The scalar result obtained from the above weighted summation is the performance degradation measurement index. The calculation formula is as follows:

[0125] ;

[0126] In the formula, This refers to the actual network request response latency collected. and These are the pre-defined lower and upper limits for network request response latency, respectively. For example, the lower limit is set at ten milliseconds, and the upper limit at one thousand milliseconds. When the network request response latency... When a network timeout is assigned the maximum penalty value, such as 9999 milliseconds, the normalized result of the network latency term will be directly locked to one after boundary truncation calculation in the formula. This accurately reflects extreme situations of severe network congestion or disconnection. (Introduction) and The boundary truncation calculation, constructed by the function, is used to ensure that when the actual probe... When drastic fluctuations occur, the normalization result is always rigidly constrained between zero and one to prevent it from going out of bounds and disturbing the overall measurement result. This represents the actual memory usage collected. The constant of the total physical memory capacity configured for each field node; The length of the queue for the actual business instructions collected; Set a threshold for the maximum allowed queue length of the program's thread pool, and Set to a natural number greater than zero to avoid triggering a division-by-zero exception. , , These are pre-defined weight coefficients, and the constraints are as follows: As a preferred approach, industrial metering sites are more sensitive to network latency anomalies, therefore, setting... , , .

[0127] S23, the weighbridge intelligent monitoring system schedules the corresponding processing logic to execute the corresponding process based on the interval range of the calculated performance degradation index.

[0128] To achieve adaptive fault tolerance and business degradation control, the weighbridge intelligent monitoring system pre-sets warning thresholds and failure thresholds in the node status evaluation thresholds of the global configuration parameters. These two specific threshold parameters are determined through offline statistical analysis of historical normal operation data and fault downtime data of the equipment.

[0129] For example, historical data of multi-dimensional operational status vectors corresponding to business lags within the past three months are selected, and this historical data is substituted into the above formula to calculate the average value of the performance degradation metric index at the time of the lag, which is used as a benchmark. In one embodiment, the warning threshold is set to 0.6, and the failure threshold is set to 0.9. The weighbridge intelligent monitoring system compares the performance degradation metric index obtained at the current moment with the above-set thresholds, and divides different status intervals based on the comparison results.

[0130] When the performance degradation metric index falls below the warning threshold, it indicates that the on-site sub-node is in a healthy operating state with sufficient resources. At this time, the weighbridge intelligent monitoring system determines that the on-site sub-node has entered the healthy state range, and then schedules the backup module 106 configured on the on-site sub-node and related business programs to execute the following routine business and backup processes.

[0131] When the performance degradation metric index rises and falls between the warning threshold and the failure threshold, it indicates that the field sub-node has entered a high-load state. This state indicates that the node is unable to process incoming business instructions in a timely manner, and the weighbridge intelligent monitoring system's scheduling throttling module 103 triggers flow limiting control and executes the subsequent sub-health state degradation process.

[0132] When the performance degradation metric index rises further and reaches the failure threshold, it indicates that the field sub-node is in a state of severe overload or communication disruption. The weighbridge intelligent monitoring system stops assigning new tasks to it and schedules the refactoring module 104 to intervene and execute the authority takeover process.

[0133] See Figure 1 and Figure 5 During the operation of the intelligent weighbridge monitoring system, when the evaluation indicators reflect that the underlying environment is in a stable state, it is necessary to ensure the continuity of daily billing operations and synchronize the historical records. Specific implementation details include the following sub-steps:

[0134] S31: When the intelligent weighbridge monitoring system is operating within a healthy range, it triggers on-site sub-nodes to perform routine weighing tasks. Specifically, it controls the gate opening via the underlying network interface device, collects vehicle positioning signals from the anti-collision radar to confirm that the vehicle is fully on the weighbridge, and reads the weight signals output by the weighbridge sensors.

[0135] In specific implementation, this step follows the aforementioned interval division logic. When the performance degradation metric index is lower than the warning threshold, the field sub-node is in a healthy state interval, and then performs routine business and completes incremental data retransmission through the backup module 106. Specifically, when the performance degradation metric index is lower than the warning threshold, the field sub-node performs routine material metering business. In this embodiment, routine material metering business refers to the data processing process in which the field sub-node controls the gate opening through the underlying network interface device, collects the vehicle positioning signal from the anti-collision radar to confirm that the vehicle is fully on the weighbridge, reads the weight signal output by the weighbridge sensor, and then combines the weighbridge sensor calibration error range value to eliminate sensor drift, and calculates the net weight of the material after tare calculation. At the same time, in the routine material metering business, the field sub-node completes the vehicle identity identification and verification according to the license plate recognition fault-tolerant matching rules, and calculates the cost of the obtained net weight of the material according to the material billing standard rate table to generate the corresponding metering bill data.

[0136] S32, the backup module 106 performs local persistence operations on the measurement results to prevent the risk of sudden data loss.

[0137] To ensure the continued availability of historical weighing records in the event of unexpected power outages or physical network disconnections, the backup module 106, configured at the field sub-nodes, records structured weighing data as a write-ahead log on the local storage device. As a preferred method, the structured weighing data is a collection of information including vehicle license plate information, gross weight, tare weight, net weight, and operation timestamps, encapsulated in a standardized format after the business program extracts the execution results of the aforementioned routine material weighing operations. Write-ahead logging is a well-known, highly reliable data protection mechanism that pre-appends the corresponding operation instructions and original data sequentially into a local, independent log file before submitting data records to the official business database. After obtaining the structured weighing data, the backup module 106 serializes it and writes it to the end of the log file. Due to physical limitations on storage space, the backup module 106 checks the remaining space capacity of the local storage device before performing the write operation. When the remaining space capacity is lower than a preset safety threshold (e.g., 5% of the remaining capacity), the backup module 106 will trigger a cyclic overwrite mechanism, automatically deleting the oldest confirmed synchronized log file with the oldest timestamp, thereby preventing program crashes caused by storage overflow. If the remaining space capacity is greater than or equal to the preset safety threshold, step S33 is triggered. In one embodiment, the local storage device is a solid-state drive or non-volatile flash memory. Because flash memory has the physical characteristic of not losing data when power is off, it ensures that even if the field node is forcibly restarted, the structured weighing data can still be reconstructed by sequentially replaying the write-ahead logs after the system recovers.

[0138] S33, through the backup module 106, a cryptographic mechanism is introduced to use a hash function to calculate and generate a tamper-proof data integrity verification digest.

[0139] To prevent the locally stored metering data from being maliciously tampered with or bit-flipped due to transmission noise during subsequent cross-network segment transmission, the backup module 106 uses a hash function to generate a data integrity verification digest. A hash function is a one-way encryption algorithm that can map input information of arbitrary length to a fixed-length string. In its implementation, the backup module 106 extracts the core numerical fields from the structured weighing data to form the base payload to be verified, and introduces an auto-incrementing sequence identifier and a random salt value to prevent replay attacks and dictionary attacks.

[0140] In this embodiment, the data integrity verification digest The specific calculation formula is as follows:

[0141] ;

[0142] In the formula, A function to calculate a secure hash algorithm with a fixed output width of 256 bits; The binary byte sequence obtained after byte stream serialization of structured weighing data; This corresponds to the auto-incrementing sequence number scalar of this record in the write-ahead log; This is a random salt constant configured within the field node, such as a 64-bit random character generated by the node during its initial power-on initialization using the media access control address of the hardware motherboard. This is a string concatenation operator. The fixed-length string calculated using the above formula is the data integrity verification digest.

[0143] S34, a non-blocking network transmission channel is established through the backup module 106 to complete the uplink synchronization of missing data.

[0144] After generating the data integrity verification digest, the backup module 106 asynchronously transmits the packaged data to the regional sub-node for incremental data replenishment. In this embodiment, asynchronous transmission means that after initiating a network transmission request, the backup module 106 hands over the data packet to an independent network thread at the underlying level. The main business program does not need to block and wait for the regional sub-node to return a receipt confirmation message, thus freeing up computing resources to process the routine material metering business of the next scheduled vehicle. This separate design can improve the throughput of the field sub-nodes.

[0145] Meanwhile, incremental data retransmission refers to the backup module 106 periodically reading the current record cursor of the locally stored write-ahead log while running in the background and comparing it with the latest received cursor issued by the regional sub-node. If the local record cursor is found to be ahead, the unsynchronized data set within the difference range between the two cursors is selected, and these incremental data, along with their corresponding data integrity verification digests, are pushed into the network transmission queue for uploading. After receiving the data packet, the regional sub-node recalculates the digest using the same hash function and a known random salt value. The validity of the data is determined by comparing whether the digests generated at both ends are strictly consistent. When the digests at both ends are consistent, the regional sub-node updates the received cursor and stores it in the database; if the digests at both ends are inconsistent, it is determined that the data packet has been damaged or tampered with during transmission. At this time, the regional sub-node will discard the abnormal data packet and issue a retransmission command with the incorrect cursor position to the backup module 106, requesting the backup module 106 to initiate a retransmission operation for the damaged data frame. Through the aforementioned closed-loop mechanism based on cursor and summary comparison, the weighbridge intelligent monitoring system achieves efficient and secure closed-loop retransmission of historical backlog data without affecting the normal measurement efficiency of the underlying system.

[0146] See appendix Figure 1 and Figure 6During the operation of the intelligent weighbridge monitoring system, when the performance degradation metric reaches the warning threshold, the throttling module 103 triggers flow control to reduce business concurrency. The specific working process includes the following sub-steps:

[0147] S41 When the intelligent monitoring system of the weighbridge determines abnormal indicators, it triggers the flow restriction intervention mechanism through the throttling module 103.

[0148] During the routine operation and monitoring of the intelligent weighbridge monitoring system, when the performance degradation metric index rises and reaches the warning threshold, it indicates that the internal computing resources of the field sub-nodes are approaching their physical capacity limit, and unprocessed data is beginning to accumulate in the pending queue. If the original request receiving frequency continues, the node may experience computational overload leading to service downtime. Based on this status quo, the throttling module 103 triggers rate limiting control. Rate limiting control can forcibly lengthen the interval between physical service events by issuing cross-level control signaling without interrupting the underlying network link.

[0149] S42, the flow control module 103 sends a flow control signal to the source gate of the field sub-node and intervenes in the dispatching process of the upstream software platform.

[0150] In step S421, the throttling module 103 sends a flow-limiting control signal to the source gate location to which the field sub-node belongs. Specifically, in order to intercept potential overload requests at the software information flow level, the throttling module 103 reads the locally configured network topology mapping table. The network topology mapping table is a structured relational data table pre-stored in a relational database management system, which records in detail the binding routing relationships between the network interconnection protocol addresses and media access control addresses of each source gate device in the factory area and its corresponding lower-level field sub-nodes. Based on the media access control address of the field sub-node currently under high load, the module traces back and locks the specific gate device that provides the initial vehicle import task. Subsequently, the throttling module 103 sends a flow-limiting control signal to the source gate location to which the field sub-node belongs. The flow-limiting control signal is a network control message with high execution priority. This message encapsulates a pause release flag to instruct the gate-end program to temporarily suspend the new vehicle entry registration and identification actions.

[0151] In step S422, after verifying the global configuration parameters, the throttling module 103 modifies the dispatching parameters of the Manufacturing Execution System (MES) to delay the allocation of the arrival time slice for reserved vehicles. Specifically, in this process, the throttling module 103 calls the data service of the MES through an application programming interface (API). The MES is a well-known software platform used within an enterprise to coordinate production scheduling, material pulling, and logistics vehicle dispatching. The throttling module 103 modifies the dispatching parameters of the MES to delay the allocation of the arrival time slice for reserved vehicles. The dispatching parameters determine the effective time window for each reserved truck to enter the factory area for weighing. Because there may be input errors in the global configuration parameters in the early stage, the throttling module 103 will first verify the failure threshold before dynamically calculating the execution time offset. With warning threshold The difference If the difference is less than or equal to the preset precision tolerance (e.g., the precision tolerance is set to 0.1), the threshold configuration is directly determined to be invalid to avoid triggering a program interruption due to a zero denominator in subsequent calculations. If the difference is determined to be greater than the preset precision tolerance (i.e., there is no abnormal input of global configuration parameters), the throttling module 103 dynamically calculates the additional time offset through an adaptive algorithm.

[0152] In this embodiment, the time offset is calculated. The specific formula is as follows:

[0153] ;

[0154] In the formula, The preset baseline delay constant, for example, is set to 30 minutes; For the natural constant An exponential function with base 0; Let be the smoothing coefficient used to control the steepness of the delayed growth curve, and the constraint condition is: As a preferred approach, a value of 1.5 is chosen. This is a performance degradation metric index calculated in real time. The warning threshold set for the preceding sequence; The failure threshold is set for the preceding sequence. Through the exponential feedback mechanism established by this formula, the closer the degradation of the field sub-node is to the failure critical point, the larger the calculated time offset value becomes. Therefore, the entry time slot allocated to subsequent vehicles will be delayed longer. After successfully calculating the time offset, the throttling module 103 submits this time offset to the manufacturing execution system, causing the manufacturing execution system to postpone the entry time slot originally allocated to subsequent queued vehicles by this time offset.

[0155] Step S423: Determine whether a network timeout or permission denial occurred when initiating the modification request. Specifically, since interface calls may fail abnormally, if the throttling module 103 encounters a network timeout or permission denial when initiating the modification request, it will directly activate the locally preset security policy and send a command with a fixed maximum delay time slice of 120 minutes to the source gate location as a fault tolerance measure. If the interface call is successful, proceed to step S43.

[0156] S43, the throttling module 103 implements control of the physical actuator to reduce the concurrent load at the end.

[0157] In addition to the peak-shaving intervention at the manufacturing execution system dispatch level, the intelligent weighbridge monitoring system also needs to implement action restrictions at the physical weighing site. The throttling module 103 generates a low-level control frame containing hardware timing parameters and sends it through to the logic programmable gate array (FPGA) inside the network interface device located at the field sub-node. Under normal conditions, when the previous vehicle finishes weighing and leaves, and the collision avoidance radar outputs a safety reset level, the barrier gate will respond to the next vehicle's request to open. Upon receiving the control frame, the throttling module 103 issues a command to increase the barrier gate opening delay interval. For example, a waiting period of up to 60 seconds is forcibly inserted into the low-level hardware control loop. This means that even if the preceding vehicle has completely emptied the weighbridge platform, the barrier gate must wait until this delay interval has elapsed before physically allowing vehicles in the queue to pass.

[0158] By implementing delayed allocation at the software dispatch level and mandatory waiting intervention at the physical site barrier level, the frequency of vehicles entering the sensor collection area is artificially reduced, thereby decreasing the concurrent traffic flowing into the site sub-nodes. This physical rate limiting and degradation method provides processing gaps for nodes under high load, allowing them to concentrate processor resources on handling backlogged tasks in their internal queues, thus alleviating business congestion caused by sudden high-frequency traffic.

[0159] See appendix Figure 1 and Figure 7 During the operation of the intelligent weighbridge monitoring system, when the node load deteriorates to the point where normal communication cannot be maintained, a bypass control strategy needs to be executed to ensure the normal operation of the weighing process. When the performance degradation metric reaches the failure threshold, the refactoring module 104 executes a takeover and issues subsequent drive commands. This specifically includes the following sub-steps:

[0160] S51 When the intelligent monitoring system of the local weighbridge determines that the node communication is blocked, it performs a forced transfer of control authority through the reconstruction module 104.

[0161] During continuous monitoring of the weighbridge intelligent monitoring system, when the performance degradation metric index further rises and reaches the failure threshold (e.g., a pre-set failure threshold of 0.9), it indicates that the processing resources of the field sub-node have been exhausted or the network link is severely blocked, making it unable to continue responding to external network requests. To prevent prolonged business process stagnation, the refactoring module 104 performs a takeover. Specifically, the refactoring module 104 generates an overwrite instruction carrying a management signature and sends this overwrite instruction to the network interface device. Upon receiving the overwrite instruction, the network interface device modifies its internal access control list register, clearing the network interconnection protocol address originally bound to the field sub-node, thereby forcibly revoking the control permissions of the field sub-node by the refactoring module 104.

[0162] In the event of an extreme network outage, if the reconstruction module 104 does not receive a response from the underlying layer within two seconds after issuing the overwrite command, it will initiate up to three retries. If all retries fail, an audible and visual alarm will be triggered and the relevant data tables will be frozen to prevent erroneous accounts.

[0163] After revoking the old control permissions and obtaining confirmation, the reconstruction module 104 requests control credentials for the corresponding peripheral device from the system's internal authentication center and transfers the exclusive control token to the regional master node. In this embodiment, this transfer operation enables the regional master node to obtain legitimate control identity for the network interface device and establishes a communication channel for direct control of the underlying hardware device.

[0164] S52 initiates queries to the underlying devices through the regional mini-node to extract physical parameters in order to construct the field environment state vector.

[0165] After completing the permission change, it is necessary to ascertain the exact status of the site due to the unexpected interruption caused by node failure. After obtaining the exclusive control token, the reconstruction module 104 sends a status query message to the network interface device. This status query message is encapsulated according to a well-known industrial control protocol and is used to instruct the network interface device to read the real-time values ​​of the hardware registers of the underlying peripherals it is connected to.

[0166] After receiving the message, the network interface device returns a set of level and digital signals containing the device's operating parameters. Based on this, the reconstruction module 104 collects the current states of the collision avoidance radar, weighbridge sensor, and barrier gate to construct a field environment state vector. As a preferred method, the collected current states include the infrared reflection level of the collision avoidance radar (e.g., 0 for no vehicle, 1 for a vehicle), the instantaneous weight digital value output by the weighbridge sensor, and the mechanical opening / closing state of the barrier gate limit switch. To eliminate the interference of dimensional differences between various features on subsequent distance calculations, the reconstruction module 104 extracts the maximum weight range value from the configuration parameters as a divisor, performs linear division mapping on the instantaneous weight digital value, and constrains it to between zero and one; while the level and switch signals retain their inherent zero or one Boolean value attributes. After the above normalization process, the reconstruction module 104 concatenates this set of data into a multi-dimensional array in a fixed order; this array is the field environment state vector.

[0167] S53, through the reconstruction module 104, uses a spatial distance comparison algorithm to determine the power outage location and restore the service drive.

[0168] Since the on-site business process is a sequence of multiple continuous actions, the reconstruction module 104 needs to determine the specific stage where the business process stalls when a fault occurs. The weighbridge intelligent monitoring system has a standard business state machine model pre-configured in its local database. This standard business state machine model includes multiple standard business stages such as vehicle waiting to enter, driving on the weighbridge, weighing stability pending confirmation, and gate opening and vehicle departure, and each business stage is associated with a baseline state vector. Combining the data obtained above, the reconstruction module 104 maps and compares the on-site environment state vectors to the business logic breakpoints.

[0169] In this embodiment, the reconstruction module 104 uses the Euclidean distance algorithm for vector comparison, and its calculation formula is as follows:

[0170] ;

[0171] In the formula, The current state vector of the field environment and the first Euclidean distance between the baseline state vectors of each business stage; This refers to the total number of feature dimensions contained in the state vector. For example, based on the data collected from the three hardware devices mentioned above—radar, weighbridge, and barrier gate—the total number of feature dimensions is... The value is 3; The first in the on-site environmental state vector Each feature component value; For the first The first reference state vector in the nth reference state vector Each feature component value is pre-defined in the database by business personnel based on standard action logic.

[0172] Using the comparison logic based on the Euclidean distance algorithm described above, the reconstruction module 104 iterates through and calculates the distance between the current state and all known reference states, and selects the calculation results. The business process containing the minimum baseline state is used as the defined business logic breakpoint.

[0173] To address potential abnormal combinations of sensor malfunctions at the site, the reconstruction module 104 sets a tolerance distance limit during comparison (e.g., a limit value of 0.15). If all calculated... All exceed the fault tolerance distance limit, or multiple different stages occur. If multiple matching conflicts occur due to the simultaneous fulfillment of the minimum value condition, the system is determined to be in an unknown abnormal state, and will stop automatic takeover and trigger a manual alarm. If a set of calculated values... If a minimum distance value less than or equal to the fault tolerance distance limit exists, and no multiple matching conflicts occur, then the business process corresponding to the baseline state where the minimum distance value is located is selected as the determined business logic breakpoint to take over from the field sub-node and send subsequent driving instructions downwards.

[0174] See appendix Figure 1 and Figure 8 In the multi-level network architecture of the intelligent weighbridge monitoring system, business scheduling under normal conditions relies on the central node for global data distribution. In industrial environments, when faced with hardware damage such as accidental cutting of the main fiber optic cable, the field communication layer will lose contact with the central management layer. To ensure that vehicles can still complete the predetermined weighing tasks within the factory area, when the central node is offline, a local area network autonomous process is initiated via broadcast module 105 to execute a local offline closed loop. The specific workflow includes the following sub-steps:

[0175] S61 When the intelligent monitoring system of the weighbridge detects a blockage in the upstream network, it confirms the offline status through the broadcast module 105 and triggers the local area network mode.

[0176] Network monitoring daemons configured within each regional sub-node periodically send application-layer status probe messages to the central node. If no response acknowledgment message is received from the central node for several consecutive detection cycles (e.g., three consecutive preset cycles), or if the underlying transmission control protocol connection in the industrial Ethernet is abnormally reset, it is determined that the backbone network is disconnected. If the central node goes offline due to network communication interruption, the broadcast module 105 initiates a local area network autonomy process. In this process, the system switches from the original vertical routing strategy of requesting business instructions from higher levels to a horizontal routing strategy of direct interaction between peer nodes, enabling each weighing area on-site to form an independent communication network that does not rely on external scheduling.

[0177] S62 establishes a low-level communication link between peer nodes through the broadcast module 105 and realizes the cross-node transfer of vehicle identity credentials.

[0178] After entering the autonomous process, the broadcast module 105 sends messages between regional sub-nodes using local communication channels. These local communication channels refer to the underlying LAN virtual channels partitioned by the plant's internal switches. These underlying LAN virtual channels enable inter-terminal communication within the same network segment without going through the core router. Since the central node cannot retrieve the vehicle's dispatch details at this time, when the vehicle passes through the regional sub-node where the origin gate is located, the broadcast module 105 of that node extracts the vehicle's physical identification (e.g., the card number read by an RFID reader when the vehicle enters the plant or the license plate sequence recognized by a camera) and generates a data digest.

[0179] In this embodiment, the broadcast module 105 horizontally delivers the hash value of the target vehicle's digital credential to the regional sub-node of the target area. To ensure the data tamper-proof characteristics during plaintext transmission within the local area network, the hash value of the digital credential is calculated using a secure hash function, and its calculation formula is as follows:

[0180] ;

[0181] In the formula, The hash value of the generated digital certificate; The function to compute a 256-bit secure hash algorithm; The physical identification string of the target vehicle; This is the entry timestamp value for the vehicle entering the source gate location; These are peer authentication key constants pre-embedded within all regional mini-nodes; Operators for string concatenation.

[0182] After completing the hash calculation, the broadcast module 105 will encapsulate the... and as auxiliary data for retrieval plain text and Plaintext data packets are horizontally delivered to all other regional nodes via LAN multicast using the User Datagram Protocol (UDP). Because LAN multicast can cause message flooding and channel congestion, the broadcast module 105 appends a unique identifier sequence number to the message header. Other nodes will discard duplicate data upon receiving a message with the same sequence number.

[0183] S63, the target node verifies the identity based on the received information and takes over subsequent independent business operations.

[0184] After a vehicle arrives at the target area, the underlying peripherals in the target area acquire the vehicle's on-site characteristics. The regional sub-node, receiving the hash value of the digital credential, compares it with its local cached characteristics and executes a local offline closed-loop process to complete the weighing operation. Specifically, when the vehicle comes to a stop on the weighbridge platform, the regional sub-node in the target area first reads the physical identity identifier collected by the underlying peripherals. Subsequently, the node searches its memory's receive buffer for previously received message records, attempting to find a corresponding entry that matches the current physical identity identifier, and extracts the entry's entry timestamp.

[0185] After obtaining the above parameters, the regional sub-node in the target area recalculates the hash value using the same hash function and key configuration, and uses this as a local cache feature for comparison. If the newly calculated hash value can be found in the received... The system finds a strictly matching record, indicating that the vehicle has a valid entry record and the flow sequence is correct. The system then executes a local offline closed loop, whereby the regional master node in the target area autonomously issues a series of control commands, such as reading the weighbridge instrument's weight value, controlling traffic light switching, and driving the barrier gate to lift.

[0186] If the corresponding physical identifier cannot be found in the local cache, or the recalculated hash value fails to match, the vehicle is determined to be an abnormal intruder or its data has been tampered with, even if it was not registered at the source gate. In this case, the regional sub-node in the target area will intercept the gate opening command, refuse to execute the weighing closed loop, and record the abnormal physical identifier and current timestamp as an anti-fraud audit log. For vehicles that match correctly, the generated metering bill data will be temporarily stored in the local storage device configured on that node, preventing further obstruction of vehicle flow while waiting for confirmation from the central node. Once the system detects that the central node is back online, the temporarily stored offline data will be merged and uploaded, ensuring the continuity of logistics metering in the factory area during network outages.

[0187] This embodiment uses a material transportation scenario in a factory area as an example to illustrate how the intelligent weighbridge monitoring system based on a multi-node collaborative architecture can achieve multi-node collaborative operation, anomaly degradation, and alarm handling under unmanned intervention. The system adopts a three-layer network and logical topology deployment, including one central node, three regional sub-nodes, and ten field sub-nodes. The bottom-layer peripherals include anti-collision radar, weighbridge sensors, and barrier gates, which are connected to the local area network via industrial Ethernet input / output modules serving as network interface devices.

[0188] 07:00 AM, System startup and parameter distribution phase:

[0189] After the system is powered on, it enters the startup phase. The initialization module 101, configured at the central node, extracts business rule elements from the relational database management system and generates global configuration parameters. Among these, the early warning threshold for node status evaluation is set to 0.6, and the failure threshold is set to 0.9.

[0190] Initialization module 101 uses the message queue bus to distribute global configuration parameters to regional sub-nodes and field nodes. Simultaneously, initialization module 101 reads the media access control address of the network interface device, combines it with the network interconnection protocol address of the field node to calculate and generate an exclusive control token, and sends it to the access control list register of the network interface device to establish access restrictions for the network interface device. For example... Figure 10 As shown, when the system just started and initialized at 07:00, the field nodes had not yet taken on external services, and their physical memory usage and network response latency were at extremely low levels, reflected in... Figure 10 The performance degradation metric index is at an extremely low baseline level. Figure 10 The vertical coordinate of the starting point shown is approximately 0.10.

[0191] 10:00 AM, normal healthy operation phase:

[0192] Logistics vehicles entered the factory area normally. The measurement module 102, configured at the regional sub-node, sent Internet Control Message Protocol (ICP-IP) probe packets to the field sub-nodes via Industrial Ethernet and read resource registers. At this time, the network request response latency was 20 milliseconds, memory usage was 30% of the total capacity, and the business instruction pending queue length was 1. The measurement module 102 constructed a multi-dimensional operating state vector from the above parameters and substituted them into a formula to calculate the performance degradation measurement index of the field sub-node. .because (0.25) is below the warning threshold (0.6). At this site, routine material weighing is performed, and sensor drift is eliminated by combining the weighbridge sensor calibration error range. The net weight of the material is then calculated after tare. Figure 10As shown, at 10:00, with the stable operation of routine material metering, the data point corresponding to this node slightly increased to 0.25. This line segment runs smoothly within the "healthy state range" below the warning threshold reference line (Y=0.6), clearly indicating that the node's hardware resources and network bandwidth are fully capable of supporting current routine business instructions. The backup module 106, configured at the field sub-node, records the structured weighing data as a write-ahead log on the local storage device and uses a hash function to generate a data integrity verification digest. Incremental data retransmission is performed to the regional sub-nodes in an asynchronous manner.

[0193] At 2:30 PM, the high-concurrency and rate-limiting control phase will begin:

[0194] The simultaneous arrival of multiple trucks led to an increase in concurrency at a certain on-site node. Data collected by the metrics module 102 showed a significant increase in the length of the pending business instructions queue, and the calculated performance degradation metric index rose to [value missing]. .because If the warning threshold (0.75) is reached and the failure threshold (0.9) is lower than the failure threshold (0.9), the throttling module 103 triggers current limiting control. Figure 10 As shown, at 14:30, due to the surge in concurrency, the slope of the broken line increased significantly, and the data point crossed the warning threshold dotted line (Y=0.6), officially entering the high-load state range defined by the warning threshold and the failure threshold (Y=0.9), thus triggering the internal scheduling mechanism of the throttling module 103. The throttling module 103 sent a flow-limiting control signal to the source gate, requesting a temporary suspension of the identification of new vehicles. Simultaneously, the throttling module 103 incorporated... Dynamically calculate the time offset This time offset is then submitted to the Manufacturing Execution System (MES), causing the MES to postpone the entry time slot originally allocated to subsequent queuing vehicles by that time offset. Please refer to... Figure 9 , Figure 9 The solid black line in the middle indicates that, as As the time slice delay increases, the delay exhibits a rapid, non-linear increase. Specifically, when the real-time acquired performance degradation metric index... When the value is 0.75, substituting the baseline delay constant of 30 minutes and the smoothing coefficient of 1.5, according to... Figure 9 The corresponding time offset can be obtained by calculating using the coordinate system projection. Approximately 63.5 minutes. After receiving this calculated value, the Manufacturing Execution System precisely postpones the entry time slice of subsequent queuing vehicles by 63.5 minutes, achieving peak shaving at the business dispatch level. In addition, the throttling module 103 issues an instruction to increase the gate opening delay interval, forcibly inserting a 60-second waiting period into the underlying hardware control loop, thereby reducing the business concurrency flowing into the field sub-nodes.

[0195] At 16:15 in the afternoon, the node failure, access control takeover, and alarm handling phase took place.

[0196] The network link at this site's sub-node experienced severe congestion. The measurement module 102 did not receive a response message within the timeout window, and therefore directly assigned the maximum penalty value to the network request-response latency. The calculated performance degradation metric index... .like Figure 10 As shown at the final time point, at 16:15, due to the penalty term assigned caused by severe network link congestion, the broken line showed a steep rise and crossed the top failure threshold horizontal line (Y=0.9), reaching... Figure 10 The extreme value of 0.95 indicates that the system has officially entered the degraded state range at the top of the chart. This data point's exceeding the limit directly terminates the aforementioned exponential rate limiting scheduling and instead activates the permission takeover process of the reconstruction module 104. Because... (0.95) When the failure threshold (0.9) is reached, the refactoring module 104 takes over the authority and generates an overwrite instruction to be sent to the network interface device to forcibly revoke the control authority of the field sub-node.

[0197] At this stage, the system exhibits two response logics:

[0198] In scenario A (successful takeover), after the refactoring module 104 issues an overwrite command and receives a confirmation, it successfully transfers the exclusive control token to the regional sub-monitor node. Subsequently, the refactoring module 104 sends status query messages to the network interface devices, collecting the current status of the underlying peripherals to construct the field environment state vector. This is achieved by calculating the Euclidean distance. The system finds that the distance between the current state and the baseline state vector (weighing stability pending confirmation) is the smallest, and sets it as the business logic breakpoint, taking over from the field sub-node to send subsequent drive instructions downwards.

[0199] For scenario B (triggered audible and visual alarms and manual alarms), if the reconstruction module 104 does not receive a response from the underlying layer within two seconds after issuing the overwrite command, and three consecutive retries fail, the reconstruction module 104 directly triggers the audible and visual alarms and freezes the relevant data tables. On the other hand, if data is successfully acquired, but the anti-collision radar is physically damaged, causing an abnormal return level, all calculated Euclidean distances will be invalidated. If all values ​​exceed the set fault tolerance distance limit (0.15), the system determines that the site is in an unknown abnormal state, stops automatic takeover, triggers a manual alarm, and notifies maintenance personnel to go to the site.

[0200] At 21:00, the backbone network is blocked and the local area network is in self-governance phase:

[0201] The main fiber optic cable in the factory area was disconnected due to external reasons. The network monitoring daemon within the regional sub-node did not receive a response confirmation message from the central node for three consecutive detection cycles, thus determining that the central node was offline. Broadcast module 105 initiated the local area network autonomy process. When a vehicle passes through the source gate location at night, broadcast module 105 at that node extracts the vehicle's physical identification identifier and calculates the hash value of the digital credential using a secure hash function. The data packet is then delivered laterally to the regional sub-node of the target area via local area network multicast using a local communication channel.

[0202] Once the vehicle reaches the target area, the underlying peripherals in the target area read the current physical identity identifier. The regional sub-node in the target area recalculates the hash value and compares it with the received identifier. A comparison is performed. If the comparison matches, it indicates that the identity verification is successful. The regional sub-node in the target area then performs a local offline closed-loop autonomous weighing operation and temporarily stores the metering bill data in the local storage device configured on that node.

[0203] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A weighbridge intelligent monitoring system based on a multi-node collaborative architecture, characterized in that, include: The initialization module is used to distribute global configuration parameters to regional sub-nodes and field sub-nodes to establish the calculation benchmark for weighbridge measurement, and to allocate exclusive control tokens to the underlying peripherals of the weighbridge to establish access restrictions; the global configuration parameters include a warning threshold and a failure threshold, and the failure threshold is greater than the warning threshold; The measurement module is used to collect the multi-dimensional operating status vector of the field sub-nodes and calculate the performance degradation measurement index of the field sub-nodes based on the multi-dimensional operating status vectors. The backup module is used to perform routine weighbridge measurement to generate weighing data according to the global configuration parameters when the performance degradation metric index is lower than the warning threshold, generate a data integrity verification summary for the weighing data, and use the data integrity verification summary to perform incremental data retransmission. The throttling module is used to trigger a flow limiting control signal when the performance degradation metric index reaches the warning threshold and is lower than the failure threshold. The refactoring module is used to revoke the control authority of the field sub-node and transfer the exclusive control token to the regional small master node to take over and issue subsequent weighbridge business driving instructions when the performance degradation metric index reaches the failure threshold. The broadcast module is used to send a message carrying the hash value of the digital certificate between the regional sub-nodes if the central node is offline, and to perform a local offline weighing closed loop based on the received message.

2. The intelligent weighbridge monitoring system based on a multi-node collaborative architecture as described in claim 1, characterized in that, The initialization module is specifically used for: An authentication signature value is generated based on the media access control address of the network interface device, the network interconnection protocol address of the field sub-node, and the current timestamp of the central node. The authentication signature value and the current timestamp are combined to form the exclusive control token and sent to the field sub-node; The authentication signature value, the current timestamp, and the network interconnection protocol address are written into the access control list register inside the network interface device for comparison and restriction, so as to allocate the exclusive control token to the underlying peripheral device.

3. The intelligent weighbridge monitoring system based on a multi-node collaborative architecture according to claim 1, characterized in that, The multidimensional runtime state vector includes network request response latency, memory usage, and business instruction pending queue length. The measurement module is specifically used for: The actual collected network request response latency is combined with the normal lower limit and the extreme upper limit value to perform extreme value standardization and boundary truncation calculations; The results of the extreme value standardization operation and the boundary truncation operation are weighted and summed with the normalized memory usage and the length of the business instruction pending queue, respectively, and the output scalar result is used as the performance degradation metric index.

4. The intelligent weighbridge monitoring system based on a multi-node collaborative architecture according to claim 1, characterized in that, The backup module is specifically used for: The weighing data is recorded as a write-ahead log on a local storage device; Extract the binary byte sequence after serializing the weighing data; The basic load is formed by the auto-incrementing sequence number in the write-ahead log and the random salt constant configured inside the field sub-nodes. A fixed-length string is generated from the underlying payload using a secure hash algorithm to serve as a data integrity verification digest.

5. The intelligent weighbridge monitoring system based on a multi-node collaborative architecture according to claim 4, characterized in that, The backup module is specifically used for: Periodically read the current record cursor of the write-ahead log; Compare the current record cursor with the latest received cursor issued by the regional sub-node; The set of unsynchronized data that falls within the difference range between the current record cursor and the latest received cursor is selected for the incremental data retransmission operation.

6. The intelligent weighbridge monitoring system based on a multi-node collaborative architecture according to claim 1, characterized in that, The throttling module is specifically used for: Send a flow control signal carrying a pause release flag to the source gate location to which the on-site sub-node belongs, so as to suspend the registration of new vehicles; Modify the dispatching parameters of the manufacturing execution system to delay the allocation of the arrival time slices of reserved vehicles using the calculated time offset; A control frame containing hardware timing parameters is generated and sent to a logic programmable gate array to increase the gate opening delay interval.

7. The intelligent weighbridge monitoring system based on a multi-node collaborative architecture according to claim 6, characterized in that, When the throttling module calculates the time offset: Calculate the difference between the current performance degradation metric index and the warning threshold, and divide it by the difference between the failure threshold and the warning threshold to obtain a quotient. The quotient is multiplied by a smoothing coefficient and then used as the exponential part of the natural constant. The time offset is obtained by multiplying the exponential portion by the reference delay constant.

8. The intelligent weighbridge monitoring system based on a multi-node collaborative architecture according to claim 1, characterized in that, The reconstruction module is specifically used for: A write command carrying a management signature is issued to clear the original address bound in the access control list register inside the network interface device, thereby revoking the control permissions of the field sub-node. The exclusive control token is transferred to the regional sub-monitor node to achieve the takeover; Send status query messages to the underlying peripheral device to collect the level and digital signals of the device operating parameters; The level and digital signals are normalized to construct a field environment state vector. The Euclidean distance algorithm is used to map the field environment state vector to the business logic breakpoint to determine the issuance position of the subsequent driving command.

9. The intelligent weighbridge monitoring system based on a multi-node collaborative architecture according to claim 1, characterized in that, The broadcast module is specifically used for: The data packet containing the hash value of the digital credential is sent to other regional mini-nodes via LAN multicast using the User Datagram Protocol. The regional sub-node that receives the data packet searches the local receiving buffer pool for the corresponding entry based on the current physical identity identifier collected by the weighbridge's underlying peripheral device. The entry timestamps in the corresponding entries are extracted, recalculated, and compared. When a match is found, the control command for the offline weighing closed loop is automatically issued.

10. A method for intelligent monitoring of weighbridges based on a multi-node collaborative architecture, characterized in that, The method is applied to the intelligent weighbridge monitoring system based on a multi-node collaborative architecture as described in any one of claims 1 to 9, and the method includes the following steps: Distribute global configuration parameters to regional sub-nodes and field sub-nodes to establish the calculation benchmark for weighbridge measurement, and assign exclusive control tokens to the underlying peripherals of the weighbridge to establish access restrictions; Collect the multi-dimensional operating status vector of the field sub-node, and calculate the performance degradation metric index of the field sub-node based on the multi-dimensional operating status vector; When the performance degradation index is lower than the warning threshold, the weighbridge performs routine measurement according to the global configuration parameters to generate weighing data, generates a data integrity verification summary for the weighing data, and uses the data integrity verification summary to perform incremental data retransmission. When the performance degradation metric reaches the warning threshold but is below the failure threshold, a flow limiting control signaling is triggered. When the performance degradation index reaches the failure threshold, the control authority of the field sub-node is revoked, and the exclusive control token is transferred to the regional sub-master node to take over and issue subsequent weighbridge business driving instructions. If the central node is offline, a message carrying the hash value of the digital certificate is sent between the regional sub-nodes, and a local offline weighing closed loop is executed based on the received message.

Images