Industrial heterogeneous network multi-protocol fusion networking and communication system

Abstract

The invention relates to the technical field of digital information transmission, and discloses an industrial heterogeneous network multi-protocol fusion networking and communication system, which comprises a heterogeneous access and semantic classification engine used for analyzing a native data identifier to determine an industrial service grade and dynamically redirecting the grade according to a system state context; the multi-dimensional dynamic queue manager is used for isolating the data frames to the virtual channel queues according to levels and monitoring the queue depth to generate priority-based flow control pause signals; according to the invention, the semantic blindness of a transmission layer is solved at a gateway entrance, and a deterministic transmission channel which is free from non-real-time data flow blocking is provided for key real-time data through mechanism coupling of industrial semantic perception and strict priority scheduling.

Description

Technical Field

[0001] This invention relates to an industrial heterogeneous network multi-protocol converged networking and communication system, belonging to the field of digital information transmission technology. Background Technology

[0002] In the current process of modern industrial automation, high-bandwidth, open Ethernet technology is gradually becoming a unified platform for carrying multiple industrial protocols, replacing traditional, closed fieldbuses and achieving deep integration of information technology and operational technology. However, this integration reveals inherent technical contradictions at the digital information transmission level: the core mechanisms of standard Ethernet switching technology, such as best-effort forwarding strategies, are not designed to be aware of the specific industrial business meaning carried within the data packets. In typical converged networks, high-bandwidth non-real-time data streams, such as machine vision image uploads or equipment maintenance log retrieval, are mixed with high-real-time periodic control data streams, such as servo motor motion commands or safety light curtain trigger signals, in the same physical transmission channel. Under such shared mechanisms, the problem of line-by-line congestion in digital information transmission is particularly prominent. When high-bandwidth data streams suddenly fill the switch's egress buffer, even if small real-time control data packets arrive, they can only queue and wait for non-critical data transmission to complete. As a result, control commands are lost or unacceptable transmission jitter occurs, causing unexpected production line shutdowns.

[0003] Existing quality of service (QoS) mechanisms largely rely on static priority configuration, essentially remaining a best-effort scheduling approach. This still cannot fundamentally prevent the risk of real-time data streams being blocked by non-real-time data streams due to the semantic blindness of the transport layer. Some technical solutions attempt to address the interconnection problem of heterogeneous networks by introducing new network architectures (such as IPv6 backbone networks) and protocol conversion devices. For example, Chinese invention patent CN112243009B discloses an IPv6-based industrial heterogeneous network multi-protocol converged networking and communication system and method, which uses a heterogeneous protocol conversion device to convert IPv4 data packets from different industrial field networks into IPv6 data packets. By using IPv6 flow tags to carry protocol type information, unified transmission on the IPv6 backbone is achieved. This solution mainly addresses the interoperability and networking issues at the network layer. When data packets are transmitted in the backbone network (such as IPv6 switches), they essentially still follow the standard Ethernet forwarding mechanism. This solution does not delve into the industrial semantics carried by the data packets and cannot distinguish whether the data is a critical control command or a non-real-time log file. When faced with the impact of high-bandwidth non-real-time data streams, it is also unable to avoid the risk of congestion and blocking caused by the semantic blindness of the transport layer. This inherent contradiction between deterministic transmission and high bandwidth utilization has become a technical problem that restricts the development of converged industrial networks.

[0004] Therefore, the technical problem to be solved by this invention is how to provide a new digital information transmission and networking method within the existing converged Ethernet architecture, so that it can have the ability to perceive industrial semantics at the entry point of transmission. Summary of the Invention

[0005] This invention provides an industrial heterogeneous network multi-protocol converged networking and communication system. Its main purpose is to solve the problem of deterministic transmission in existing converged networks, where critical real-time data streams are blocked by non-real-time bandwidth data streams due to the lack of industrial semantic awareness capabilities at the transport layer.

[0006] To achieve the above objectives, the present invention provides an industrial heterogeneous network multi-protocol converged networking and communication system, the system comprising:

[0007] The heterogeneous access and semantic classification engine is used to receive native data frames from multiple source ports and map the native data identifiers in the native data frames to the industrial service level by referring to the semantic classification rule base. The heterogeneous access and semantic classification engine is also used to monitor native data frames mapped to high priority to update the system state context according to the content, and to check the system state context to dynamically redirect the industrial service level of subsequent native data frames.

[0008] The multidimensional dynamic queue manager includes multiple independent virtual channel queues corresponding to industrial service levels. The multidimensional dynamic queue manager is also used to store native data frames into the corresponding virtual channel queues according to industrial service levels, monitor the queue depth of the virtual channel queues corresponding to non-real-time industrial service levels, and generate a priority-based flow control pause signal when the queue depth exceeds a preset high water level, and send the priority-based flow control pause signal to the source port corresponding to the native data frame.

[0009] A non-blocking egress scheduler is used to connect multiple independent virtual channel queues and uplink network ports, and select data frames from multiple independent virtual channel queues for transmission according to a preset strict priority of industrial service level.

[0010] Preferably, the semantic classification rule base also includes dynamic redirection logic, which defines the association between triggering conditions and priority redirection actions. The heterogeneous access and semantic classification engine are specifically used to monitor the content of native data frames mapped to high priority to determine whether the system state context meets the triggering conditions, and to execute priority redirection actions to change the industrial service level of subsequent native data frames.

[0011] Preferably, the heterogeneous access and semantic classification engine is also used to map industrial service levels to data priorities of network transmission standards; the priority-based flow control pause signal is a flow control signal based on data priority, which only indicates the pause of data priorities corresponding to non-real-time industrial service levels, without affecting the transmission of other data priorities.

[0012] Preferably, the non-blocking egress scheduler is specifically used to extract data frames from the next lower priority virtual channel queues only when one or more highest priority virtual channel queues are empty; wherein, the highest priority virtual channel queue is the virtual channel queue corresponding to the highest priority industrial service level, and the next lower priority virtual channel queue is the virtual channel queue corresponding to the next lower priority industrial service level.

[0013] Preferably, the non-blocking egress scheduler is also used to provide a minimum bandwidth guarantee for virtual channel queues corresponding to non-real-time industrial service levels; the minimum bandwidth guarantee is implemented through a token bucket algorithm, wherein the token bucket algorithm replenishes the virtual channel queues corresponding to non-real-time industrial service levels with tokens within a preset time period, satisfying the following: ,in, For token quantity, The preset minimum guaranteed rate is for minimum bandwidth. This is a preset time period.

[0014] Preferably, the system further includes: a queue statistics module connected in parallel with a multi-dimensional dynamic queue manager; the queue statistics module is used to collect queue operation statistics of multiple independent virtual channel queues, the queue operation statistics include queue depth, enqueue rate and packet loss count, and output diagnostic indication information to characterize the network operation health status based on the deviation of the queue operation statistics from a preset baseline.

[0015] Preferably, the multidimensional dynamic queue manager is also used to generate a flow control recovery signal when the queue depth is lower than a preset low water level, and send the flow control recovery signal to the source port to instruct the source port to resume data transmission.

[0016] Preferably, heterogeneous access and semantic classification engines connect to heterogeneous industrial networks; native data identifiers include CANID, Modbus register address, and Profinet frame identifier.

[0017] Preferably, industrial service levels include: critical real-time levels for safety interlocks and emergency shutdowns, cyclic real-time levels for servo commands and periodic I / O, and non-real-time bandwidth levels for visual images and log files.

[0018] Preferably, the semantic classification rule base is dynamically configurable, allowing external monitoring and data acquisition SCADA and manufacturing execution system MES to modify the mapping relationship between native data identifiers and industrial service levels at runtime.

[0019] Compared with the prior art, the beneficial effects of the present invention are:

[0020] 1. Through industrial semantic parsing performed at the gateway entry point by heterogeneous access and semantic classification engine, the original data frames (such as CANID or Modbus addresses) that are originally transparent to the transmission system and different industrial networks are translated and marked as a unified industrial service level. This level label directly drives the multi-dimensional dynamic queue manager to store data streams of different levels into mutually logically isolated virtual channel queues. This mechanism of semantic awareness and isolated queuing coupling enables the subsequent non-blocking exit scheduler to perform strict priority scheduling based on this level. Thus, on the converged digital information transmission link, a predictable logical priority channel is opened for the critical real-time control data stream, so that it is not affected by the transmission blockage and jitter caused by the sudden congestion of non-real-time, high-bandwidth data streams (such as images or logs).

[0021] 2. By adding dynamic redirection logic to the semantic classification engine, dynamic matching of transmission priority with industrial operating conditions is achieved. This mechanism monitors the content of high-priority data streams to update the system state context. Using this context as a trigger, under specific operating conditions such as alarms, the industrial service level of the associated data (such as conventional temperature parameters) is dynamically redirected from non-real-time to critical real-time. Within the transmission system, the self-consistent mechanism of using data streams to control classification streams ensures that diagnostic data will inevitably obtain the highest transmission priority when the system most needs diagnostic data, avoiding the risk of untimely diagnosis due to the blockage of conventional non-real-time traffic for critical diagnostic information.

[0022] 3. This invention constructs a closed-loop transmission mechanism that balances real-time determinism with bandwidth data integrity. By utilizing queue depth monitoring of a multi-dimensional dynamic queue manager, the congestion state of non-real-time queues is transformed into a flow control pause signal. Through a semantic classification engine, internal industrial service levels are mapped to standard transmission data priorities, making the pause signal a priority-based precise control signal. This synergy of queue status monitoring, priority mapping, and priority flow control not only avoids buffer overflows and data loss of non-real-time data streams at the gateway entry point, but also prevents the pause signal from mistakenly blocking high-priority real-time data streams such as servo commands sharing the same port in complex network conditions with shared ports. This achieves deterministic and integrity transmission of multiple service streams in converged networks. Attached Figure Description

[0023] Figure 1 This is a system functional architecture diagram of semantic classification and priority scheduling of the present invention;

[0024] Figure 2 This is a comparison chart of the real-time and non-real-time queue depth characteristics of the present invention.

[0025] Figure 3 This is a diagram showing the interactive deployment of the system of the present invention with MES / SCADA and field equipment. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the invention and are not intended to limit the invention.

[0027] This invention discloses an industrial heterogeneous network multi-protocol converged networking and communication system. The system's architecture mainly includes a heterogeneous access and semantic classification engine, a multi-dimensional dynamic queue manager, and a non-blocking egress scheduler. The heterogeneous access and semantic classification engine receives and parses the raw data frames from the heterogeneous network, mapping them to a unified industrial service level. The multi-dimensional dynamic queue manager stores the data frames into mutually isolated virtual channel queues according to this level. The non-blocking egress scheduler extracts data frames from each queue and sends them to the uplink network port according to a strict priority strategy. The heterogeneous access and semantic classification engine includes multiple source ports for physically connecting to heterogeneous industrial networks, such as CAN bus, Modbus RTU / TCP network, or Profinet network. This engine is used to parse raw data frames at the data inflow entry point to identify the industrial semantics they carry. The engine parses the raw data identifier within the raw data frame, which may include CANID, Modbus register address, or Profinet frame identifier, etc., used to distinguish data types. The engine maintains a semantic classification rule base internally. This rule base is a configurable mapping table that establishes the correspondence between native data identifiers and industrial service levels. Industrial service levels are internal classifications based on the criticality of industrial operations. For example, these classifications may include: critical real-time level (ICS-1) for safety interlocks and emergency shutdowns, periodic real-time level (ICS-2) for servo commands and periodic I / O, and non-real-time bandwidth level (ICS-3) for visual images and log files. When the engine's parsing unit receives a native data frame (taking a frame with CANID 0x1A5 as an example), it extracts the identifier 0x1A5 and queries the semantic classification rule base. If the rule base defines CANID 0x1A5 as corresponding to ICS-2, the data frame is marked as a periodic real-time level. This semantic classification rule base is dynamically configurable, allowing external Supervisory Control and Data Acquisition (SCADA) or Manufacturing Execution System (MES) to modify the mapping relationship between native data identifiers and industrial service levels through standard communication interfaces during system operation.

[0028] The heterogeneous access and semantic classification engine is also used to monitor native data frames mapped to high priority (such as ICS-1), updating the local system state context based on their content. The system state context (e.g., status flags) is used to characterize the current industrial operating conditions. The semantic classification rule base further includes dynamic redirection logic, which defines the association between triggering conditions and priority redirection actions. The triggering conditions are associated with changes in the system state context. Taking the reactor monitoring scenario as an example, the emergency pressure relief alarm frame (CANID 0x001) is statically defined as ICS-1 in the rule base; the reactor bottom temperature data (Modbus address 40001) is statically defined as ICS-3. The dynamic redirection logic can be defined as follows: Meaning: When the system state context is updated to alarm state upon receiving CANID0x001, a priority redirection action is executed. All subsequent data frames with Modbus address 40001 are dynamically redirected to ICS-1 industrial service level. The engine executes the priority redirection action after detecting that the system state context meets this trigger condition, dynamically changing the industrial service level of subsequent native data frames. A multi-dimensional dynamic queue manager internally sets up multiple independent logically isolated virtual channel queues. Each virtual channel queue corresponds to an industrial service level such as ICS-1, ICS-2, and ICS-3. The classification engine stores data frames marked with their service level into their corresponding virtual channel queue. The multi-dimensional dynamic queue manager also monitors the queue depth of virtual channel queues corresponding to non-real-time Industrial Service Level (ICS-3) as an example, and sets a preset high-water mark for the queue. This preset high-water mark can be determined based on the physical buffer size of the queue, such as setting it to 80% of the buffer capacity. When the queue depth exceeds this preset high-water mark, the manager generates a priority-based flow control pause signal. The heterogeneous access and semantic classification engine also maps Industrial Service Level (ICS-3) as an example to data priorities in network transmission standards, such as Prio2 in IEEE 802.1p. The priority-based flow control pause signal is a flow control signal based on this data priority, using IE... Taking the EE802.1Qbb Priority Flow Control (PFC) PAUSE frame as an example, this PFC frame contains a priority mask that indicates only the pause of data priority corresponding to non-real-time Industrial Service Level (ICS-3, Prio2), without affecting the transmission of other data priorities (such as Prio4 corresponding to ICS-2). The Multidimensional Dynamic Queue Manager sends this PFC PAUSE frame to the source port corresponding to the non-real-time data frame. The Multidimensional Dynamic Queue Manager also sets a preset low watermark for the queue, such as 20% of the queue's physical buffer. When the queue depth is lower than the preset low watermark, the manager generates a flow control recovery signal and sends the flow control recovery signal to the source port.

[0029] The non-blocking egress scheduler connects multiple independent virtual channel queues and uplink network ports such as Gigabit Ethernet ports. The scheduler executes scheduling according to a preset strict priority (SP) for the Industrial Service Level. Specifically, the non-blocking egress scheduler checks the highest priority virtual channel queue; if the queue is not empty, it extracts data frames for transmission. Only when the highest priority virtual channel queue is empty is data frames extracted from the next lower priority virtual channel queue, and so on. The current priority queue is processed only when all higher priority queues are empty. The non-blocking egress scheduler also provides a minimum bandwidth guarantee for virtual channel queues corresponding to non-real-time Industrial Service Levels. This minimum bandwidth guarantee is implemented using a token bucket algorithm, with the system executing according to a preset time period. For example, replenish the token quantity of this queue every 1ms. The token quantity satisfies: ,in, The token quantity is in bytes. The default minimum guaranteed rate, such as 50Mbps, is set to ensure minimum bandwidth. For a preset time period, when the non-blocking egress scheduler schedules the non-real-time virtual channel queue, it is only allowed to send data frames if there are enough tokens in the token bucket of the queue, and the corresponding tokens are deducted after sending. This system may also include a queue statistics module, which is connected in parallel with the multi-dimensional dynamic queue manager. It is used to periodically collect the queue operation statistics of multiple independent virtual channel queues, such as every 100ms. The queue operation statistics include the real-time queue depth, peak queue depth, enqueue rate in packets / second or bytes / second, and packet loss count for each virtual channel queue. This is used to establish a preset baseline. By collecting the queue operation statistics under normal operating conditions, the enqueue rate of ICS-2 is stabilized at 1000±50pps. During system operation, the real-time collected queue operation statistics are continuously compared with the preset baseline. When the deviation of the queue operation statistics from the preset baseline exceeds a preset threshold, the module outputs diagnostic indication information to characterize the network's operational health status. This information can clearly indicate an anomaly in the periodic real-time stream ICS-2.

[0030] Example 1: This example demonstrates a specific application of a typical industrial converged network digital information transmission conflict scenario. On a converged industrial automated production line, a high-bandwidth machine vision quality inspection system and a safety light curtain device used for safety interlocking share the same uplink Ethernet link to access the communication system of this invention. At a certain moment, the vision system begins uploading a 10MB high-definition image file, generating a dense, high-bandwidth data stream; almost simultaneously, the operator triggers the safety light curtain, generating a tiny safety stop instruction data frame containing only a few bytes; both data streams arrive simultaneously at the system's heterogeneous access and semantic classification engine; this engine... Based on its semantic classification rule base, the raw data identifiers of a large number of image data frames of the vision system are determined to be non-real-time bandwidth level (ICS-3), and the raw data identifiers of the safety grating are determined to be critical real-time level (ICS-1). The multi-dimensional dynamic queue manager stores all ICS-3 image data frames into their corresponding virtual channel queue VCQ-NRT-BW, and stores the ICS-1 safety instruction frame into a completely independent logically isolated virtual channel queue VCQ-CRT. The VCQ-NRT-BW queue has a large backlog of data to be sent, while the VCQ-CRT queue only contains data frames.

[0031] The non-blocking exit scheduler starts working and checks the highest priority VCQ-CRT according to the preset strict priority. The scheduler finds that the queue is not empty and immediately extracts the ICS-1 safety instruction frame at the head of the queue and sends it to the uplink network port. At this time, all data frames in the VCQ-NRT-BW queue are not served. The critical ICS-1 safety instruction frame is transmitted with minimal delay to ensure production line safety. After the ICS-1 data frame is sent, the scheduler checks that the high priority queue is empty and then begins to extract and send non-real-time ICS-3 image data frames from the VCQ-NRT-BW. No critical data blockage occurs during the entire process.

[0032] Example 2: This example objectively verifies the deterministic guarantee effect of the technical solution of the present invention on the transmission of critical real-time data streams under mixed traffic loads. A test platform simulating an industrial converged network was built. This platform includes a periodic real-time (ICS-2) traffic generator, a non-real-time bandwidth (ICS-3) traffic generator, and a high-precision network analyzer for capturing and measuring the end-to-end latency of data frames. The ICS-2 traffic generator is used to simulate servo commands, set to send 64-byte data frames at a constant rate with a period of 1ms. The ICS-3 traffic generator is used to simulate visual image or log uploads, sending 1500-byte data frames at an adjustable rate to generate different background bandwidth loads. Two sets of experiments were set up: a control group and the sample group of the present invention. The control group uses the IEEE 802.1p standard. In an industrial Ethernet switch with QoS (Quality of Service), ICS-2 traffic is mapped to high priority (Prio6), and ICS-3 traffic is mapped to low priority (Prio1). The sample group of this invention uses a communication system as described in the specific implementation. Its heterogeneous access and semantic classification engine maps ICS-2 traffic to periodic real-time levels and stores it in the VCQ-PRT queue, and maps ICS-3 traffic to non-real-time bandwidth levels and stores it in the VCQ-NRT-BW queue. Strict priority scheduling is performed by a non-blocking egress scheduler. Under the condition of uplink network port bandwidth of 1Gbps, the maximum end-to-end delay and jitter of ICS-2 periodic real-time data frames were tested in both the control group and the sample group of this invention as the ICS-3 background traffic load gradually increased from 0% to 95% link occupancy. The test data are recorded in Table 1.

[0033] Table 1: Comparison of Key Data Transmission Latency under Different Non-Real-Time Traffic Loads

[0034]

[0035] As shown in Table 1, the control group exhibited lower latency under no-load conditions. With increasing ICS-3 background traffic load, the maximum latency and jitter of the ICS-2 data frames increased sharply. When the load reached 95%, the maximum latency reached 24.8ms, which caused transmission failure for servo control with a 1ms cycle. This was due to the limitations of the standard QoS mechanism and outgoing buffer congestion. Data from the present invention's sample group showed that regardless of whether the ICS-3 background traffic load increased from 0% to 95%, the maximum latency of its ICS-2 data frames remained below 0.2ms, and the latency jitter remained at an extremely low level of 0.06ms. This data objectively confirms that the system of the present invention, through semantic awareness, queue isolation, and strict priority scheduling mechanisms, protects critical real-time data streams from congestion caused by non-real-time bandwidth data streams at the digital information transmission level.

[0036] Example 3: This example combines Figures 1 to 3 This describes a multi-protocol converged networking and communication system for heterogeneous industrial networks, such as... Figure 1 As shown, native data frames from heterogeneous industrial networks such as CAN, Modbus, and Profinet source ports enter the heterogeneous access and semantic classification engine. This engine is responsible for parsing the native identifier and mapping the industrial service level. It contains a semantic classification rule base and dynamic redirection logic. The processed and generated classified data frames with industrial service levels are sent to the multi-dimensional dynamic queue manager. The manager isolates the data into virtual channel queues according to the level and monitors the queue depth. When needed, the manager can send a priority-based flow control pause signal back to the heterogeneous industrial network source port. Scheduled data frames are sent from the manager to the non-blocking egress scheduler to perform strict priority scheduling to ensure real-time flow. The minimum bandwidth guarantee can be selected. Finally, it connects to the uplink backbone network through the uplink network port. The queue statistics module is connected in parallel with the multi-dimensional dynamic queue manager to collect queue operation statistics, perform network health diagnosis, and output diagnostic indication information to external monitoring HMI / SCADA.

[0037] like Figure 2 As shown in the figure, the vertical axis represents the queue depth (number of data frames), and the horizontal axis represents time (seconds). The figure displays three curves: the queue depth of the critical real-time level ICS-1 remains consistently at an extremely low level close to zero; the queue depth of the periodic real-time level ICS-2 remains at a low level with minimal fluctuations; while the queue depth of the non-real-time bandwidth level ICS-3 fluctuates, with its number of data frames being significantly higher than the other two levels, peaking at 15 seconds. Figure 3 As shown, the heterogeneous industrial network fieldbus / industrial Ethernet at the bottom layer connects various devices, including CAN bus devices such as safety light curtains, Modbus devices such as reactor sensors, Profinet devices such as servo controllers, and high-bandwidth devices such as vision cameras. The native data frames generated by these devices are connected to the core gateway of the industrial heterogeneous network multi-protocol converged networking and communication system through the source port. This core gateway is connected to the uplink backbone network such as industrial Ethernet through the uplink network port. The upper-layer manufacturing execution system (MES) can dynamically configure the system rules through this backbone network. The monitoring and data acquisition SCADA system is also connected to the backbone network to transmit network health diagnosis information to the human-machine interface (HMI).

[0038] Example 4: The process for determining the preset high watermark (HWM) and preset low watermark (LWM) of the non-real-time virtual channel queue in the multi-dimensional dynamic queue manager is as follows: Determine the system's link round-trip time (RTT) and set it to... Determine the maximum burst rate of the non-real-time data stream. When the 1Gbps port is used for visual image uploads, this rate is 1Gbps. Based on this, calculate the maximum amount of additional burst data the system still needs to receive within the RTT time. If the total physical buffer size of the virtual channel queue is 4 Mbits, then the default high watermark (HWM) is set to [value missing]. The preset low water level (LWM) setting aims to avoid frequent oscillations in the flow control signal. Its value is separated from the high water level (HWM) by a hysteresis window. This window size is set to be greater than the maximum data frame MTU of 1500 bytes. Therefore, LWM can be set to HWM minus two MTUs. The specific implementation of the dynamic redirection logic in the heterogeneous access and semantic classification engine is as follows: The system state context is implemented as a 32-bit state flag register; each bit in this register is associated with a specific industrial condition, such as a reactor overpressure alarm or a robotic arm A-axis fault; when the engine detects an alarm frame defined as ICS-1 with CANID 0x001 arriving, the internal logic can use a lookup table to map the CANID to bit 5 of the state flag register and set that bit to 1, thereby updating the system state context; the algorithmic judgment process of the engine when performing semantic classification is as follows: The first step is to parse a native data frame with Modbus address 40001 as usual and obtain its default level ICS-3 from the semantic classification rule base. The second step is to check whether the entry in the rule base is associated with dynamic redirection logic. The third step is to check whether bit 5 of the status flag register is 1 if an association exists. The fourth step is to check whether bit 5 of the status flag register is 1 if bit 5 is 1, indicating that the reactor overpressure alarm is activated. The engine will then override the default level, dynamically redirect the industrial service level of the data frame to ICS-1, and send it to the virtual channel queue corresponding to ICS-1.

[0039] The preset minimum guaranteed rate for minimum bandwidth guarantee in the non-blocking egress scheduler The determination process is as follows: Determine the total uplink network port bandwidth of the system. Set to 1000Mbps; calculate the theoretical peak bandwidth of all real-time streams defined as high priority (ICS-1 and ICS-2 in this case). The peak bandwidth is calculated by dividing the size of each periodic data frame by its transmission period; for an ICS-2 servo command with a data frame size of 100 bytes and a transmission period of 1ms, the bandwidth is... Assume the sum of the peak bandwidths of all ICS-1 and ICS-2. 150Mbps; It can be set in the following ways: in, This represents the total bandwidth of the uplink network ports. This represents the sum of the theoretical peak bandwidths of all high-priority real-time streams, with 0.8 serving as a safety margin factor. Based on the above settings, ; The value must also meet the minimum data integrity requirements of non-real-time applications. If the minimum requirements of visual applications (set to 50Mbps) are lower than the above calculated value, then the calculated value shall be used. If they are higher, then the allocation of real-time services or the total bandwidth shall be re-evaluated or upgraded.

[0040] Example 5: This example, when first deployed in a new industrial site, is used to construct a semantic classification rule base and standardize the baseline of the calibration queue statistics module. When the system initially connects to the production line network, the semantic classification rule base is empty, and the system enters rule learning mode. In this mode, the operator sequentially activates various heterogeneous devices on the production line, causing them to generate characteristic data streams. When a raw data frame with CANID 0x001 from a safety light curtain arrives for the first time, the heterogeneous access and semantic classification engine captures this undefined identifier and prompts the operator to assign it an industrial service level; the operator assigns it a critical real-time level. When a periodic data frame with CANID 0x1A5 from a servo controller arrives, the engine captures this new identifier, and the operator assigns it a periodic real-time level. When an image data stream from a vision camera arrives, the engine captures an identifier, which can be TCP port number 9. 100, the operator assigns it to a non-real-time bandwidth level; the system stores these mapping relationships from native data identifiers to industrial service levels into the semantic classification rule base until all key data flows of the production line are defined; after the semantic classification rule base is filled, the system starts the baseline calibration mode; the operator puts the production line into a standard, stable operating condition and continuously runs a preset calibration cycle, which is set to 600 seconds; during this cycle, the queue statistics module continuously collects the queue operation statistics of all classified virtual channel queues, especially the inbound rate of the queue corresponding to the real-time level of the cycle; after the calibration cycle ends, the module calculates the statistical average value of the inbound rate of the queue during the cycle, such as 998.5pps and the standard deviation, such as 4.3pps; the system then stores the statistical average value and the fluctuation range based on the standard deviation, such as the average value ± 3 times the standard deviation, as the preset baseline for subsequent network health status diagnosis.

[0041] Example 6: When the communication system needs to interact with the upper-level Manufacturing Execution System (MES) for online updates of the semantic classification rule base, standardized engineering procedures are executed. A rule update interface is provided using the reserved Modbus TCP register address space, such as 40100 to 40199. When the MES needs to dynamically modify a classification rule, the original data identifier of the new rule (e.g., CANID0x3F0) and the new industrial service level code (e.g., changing from ICS-3 to ICS-2) are written to the temporary storage area of ​​this interface, such as addresses 40100 and 40101. After the data is written, the MES writes a value of 1 to the commit flag register, such as 40199. The heterogeneous access and semantic classification engine periodically scans this flag. When a value of 1 is detected, the engine locks the rule base, performs an atomic operation, overwrites the data in the temporary storage area into the semantic classification rule base, and then clears the flag. When the local Human-Machine Interface (HMI)... When real-time monitoring of the system's network health status is required, the diagnostic indication information of the queue statistics module is provided in the following way: The diagnostic indication information is provided in the form of a set of read-only data registers, which can be periodically read by an external HMI or SCADA system; This data structure allocates an independent statistical data area for each virtual channel queue, such as VCQ-CRT, VCQ-PRT, etc.; Taking the area corresponding to VCQ-PRT as an example, it contains: current queue depth, 16-bit integer; peak queue depth, 16-bit integer; average enqueue rate, 32-bit floating-point number, unit pps; and baseline deviation status word, 8-bit integer; The baseline deviation status word is calculated in real time by the queue statistics module based on a preset baseline, where bit 0 is 1 to indicate an abnormally high rate, and bit 1 is 1 to indicate a persistently high depth; The external HMI reads this status word and directly displays the periodic real-time stream ICS-2 alarm without parsing the original statistical data.

[0042] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0043] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. An industrial heterogeneous network multi-protocol converged networking and communication system, characterized in that, The system includes: The heterogeneous access and semantic classification engine is used to receive native data frames from multiple source ports and map the native data identifiers in the native data frames to the industrial service level by referring to the semantic classification rule base. The heterogeneous access and semantic classification engine is also used to monitor native data frames mapped to high priority to update the system state context according to the content, and to check the system state context to dynamically redirect the industrial service level of subsequent native data frames. The multidimensional dynamic queue manager includes multiple independent virtual channel queues corresponding to industrial service levels. The multidimensional dynamic queue manager is also used to store native data frames into the corresponding virtual channel queues according to industrial service levels, monitor the queue depth of the virtual channel queues corresponding to non-real-time industrial service levels, and generate a priority-based flow control pause signal when the queue depth exceeds a preset high water level, and send the priority-based flow control pause signal to the source port corresponding to the native data frame. A non-blocking egress scheduler is used to connect multiple independent virtual channel queues and uplink network ports, and select data frames from multiple independent virtual channel queues for transmission according to a preset strict priority of industrial service level.

2. The industrial heterogeneous network multi-protocol converged networking and communication system according to claim 1, characterized in that, The semantic classification rule base also includes dynamic redirection logic, which defines the association between triggering conditions and priority redirection actions. The heterogeneous access and semantic classification engine are specifically used to monitor the content of native data frames mapped to high priority to determine whether the system state context meets the triggering conditions, and to execute priority redirection actions to change the industrial service level of subsequent native data frames.

3. The industrial heterogeneous network multi-protocol converged networking and communication system according to claim 1, characterized in that, The heterogeneous access and semantic classification engine is also used to map industrial service levels to data priorities in network transmission standards; the priority-based flow control pause signal is a flow control signal based on data priority, which only indicates the pause of data priorities corresponding to non-real-time industrial service levels, without affecting the transmission of other data priorities.

4. The industrial heterogeneous network multi-protocol converged networking and communication system according to claim 1, characterized in that, The non-blocking egress scheduler is specifically used to extract data frames from the next lower priority virtual channel queues only when one or more of the highest priority virtual channel queues are empty; wherein, the highest priority virtual channel queues are the virtual channel queues corresponding to the highest priority industrial service level, and the next lower priority virtual channel queues are the virtual channel queues corresponding to the next lower priority industrial service level.

5. The industrial heterogeneous network multi-protocol converged networking and communication system according to claim 4, characterized in that, The non-blocking egress scheduler is also used to provide minimum bandwidth guarantees for virtual channel queues corresponding to non-real-time industrial service levels. These minimum bandwidth guarantees are implemented using a token bucket algorithm, which replenishes the virtual channel queues corresponding to non-real-time industrial service levels with tokens within a preset time period, satisfying the following conditions: ,in, For token quantity, The preset minimum guaranteed rate is for minimum bandwidth. This is a preset time period.

6. The industrial heterogeneous network multi-protocol converged networking and communication system according to claim 1, characterized in that, The system also includes: a queue statistics module connected in parallel with a multi-dimensional dynamic queue manager; the queue statistics module is used to collect queue operation statistics of multiple independent virtual channel queues, including queue depth, enqueue rate and packet loss count, and outputs diagnostic indication information to characterize the network operation health status based on the deviation of the queue operation statistics from the preset baseline.

7. The industrial heterogeneous network multi-protocol converged networking and communication system according to claim 3, characterized in that, The multidimensional dynamic queue manager is also used to generate a flow control recovery signal when the queue depth is lower than a preset low water level, and send the flow control recovery signal to the source port to instruct the source port to resume data transmission.

8. The industrial heterogeneous network multi-protocol converged networking and communication system according to claim 1, characterized in that, Heterogeneous access and semantic classification engine connect heterogeneous industrial networks; native data identifiers include CANID, Modbus register address, and Profinet frame identifier.

9. The industrial heterogeneous network multi-protocol converged networking and communication system according to claim 1, characterized in that, Industrial service levels include: critical real-time levels for safety interlocks and emergency shutdowns, cyclic real-time levels for servo commands and periodic I / O, and non-real-time bandwidth levels for visual images and log files.

10. The industrial heterogeneous network multi-protocol converged networking and communication system according to claim 1, characterized in that, The semantic classification rule base is dynamically configurable, allowing external monitoring and data acquisition SCADA and manufacturing execution system MES to modify the mapping relationship between native data identifiers and industrial service levels at runtime.

Images