Methods and systems for physical interaction data processing of power grid equipment

Abstract

This application provides a method and system for processing physical interaction data of power grid equipment. The method constructs flow description information for each periodic data stream based on the message transmission period and message length of each intelligent electronic device (IED) in the process layer. Based on the flow description information, worst-case arrival analysis is performed for each output port in the TSN. Based on the results of the worst-case arrival analysis, a gating schedule for each output port in the TSN is generated or adjusted. Based on the generated gating schedule, a transmission phase offset parameter is assigned to each IED. During the operation of the TSN network, delay measurement is performed on the periodic data streams to obtain delay statistics. Based on the delay statistics, the gating schedule is adaptively adjusted to maintain message transmission delay jitter within a preset upper threshold, thereby improving the determinism and reliability of critical periodic data stream transmission in the process layer of the intelligent substation.

Description

Technical Field

[0001] This application relates to data processing technology, and more particularly to a method and system for processing physical interaction data of power grid equipment. Background Technology

[0002] Time-Sensitive Networking (TSN) technology introduces a deterministic latency mechanism on top of standard Ethernet, providing reliable assurance for the transmission of critical periodic data streams such as sampled values ​​and substation events oriented towards general objects in the process layer of smart substations. One of its core technologies, the Time-AwareShaper (TAS), reserves dedicated transmission time slots by configuring gate control lists (GCLs) for different priority queues, thereby ensuring low-latency transmission of high-priority messages.

[0003] However, in practical engineering configurations, gating schedules are often based on simple estimates of the average bandwidth of each data stream. This approach fails to adequately consider the periodic data streams sent by multiple intelligent electronic devices in the process layer network, whose transmission periods and initial phases can have various combinations. In the worst-case phase combination, packets from multiple data streams may arrive at the same output port of the switch within the same time window, causing the instantaneous data volume at that port to exceed the transmission capacity of the reserved gating time slot. In this case, the excess packets will be forced to queue and wait for the next transmission opportunity, thus introducing periodic delay jitter into the originally designed deterministic transmission process.

[0004] This periodic delay jitter, when transmitted to protection, measurement and control devices, will be converted into time alignment error of sampled values ​​or time delay randomness of protection action commands. This directly affects the accuracy and reliability of synchronous sampling, phasor calculation and relay protection actions, and weakens the determinism and predictability of advanced application functions of smart substations. Summary of the Invention

[0005] This application provides a method and system for processing physical interaction data of power grid equipment, which solves the technical problem that the existing technology, when configuring the TSN network gating schedule, only considers the average bandwidth and ignores the worst-case phase alignment, resulting in delay jitter in the critical periodic data stream. It realizes precise control and optimization of message transmission delay and jitter.

[0006] In a first aspect, this application provides a method for processing physical interaction data of power grid equipment, applied to the process layer of a smart substation employing Time-Sensitive Networking (TSN); the method includes:

[0007] Based on the message transmission period and message length of each intelligent electronic device (IED) in the process layer, the flow description information of each periodic data stream is constructed.

[0008] Based on the flow description information, worst-case arrival analysis is performed for each output port in the TSN;

[0009] Based on the results of the worst-case arrival analysis, generate or adjust the gating schedule for each output port in the TSN;

[0010] Based on the generated gating schedule, a transmission phase offset parameter is assigned to each IED to perform network-level phase orchestration of the transmission times of the periodic data stream;

[0011] During the operation of the TSN network, delay is measured on the periodic data stream to obtain delay statistics.

[0012] Based on the aforementioned delay statistics, the gating schedule is adaptively adjusted.

[0013] Secondly, this application provides a power grid equipment physical interaction data processing system, comprising:

[0014] The module is used to construct the flow description information of each periodic data stream based on the message transmission period and message length of each intelligent electronic device (IED) in the process layer of a smart substation using Time-Sensitive Networking (TSN).

[0015] The execution module is used to perform worst-case arrival analysis for each output port in the TSN based on the flow description information;

[0016] The generation module is used to generate or adjust the gating schedule for each output port in the TSN based on the results of the worst-case arrival analysis.

[0017] The transmitting module is used to assign a transmitting phase offset parameter to each of the IEDs based on the generated gating schedule;

[0018] The measurement module is used to measure the delay of the periodic data stream during the operation of the TSN network and obtain delay statistics.

[0019] The adjustment module is used to adaptively adjust the gating schedule based on the delay statistics.

[0020] The power grid equipment physical interaction data processing method and system provided in this application generates a gating schedule by performing worst-case arrival analysis. This ensures that, under any message phase combination, the transmission capacity of each gating time slot is sufficient to cover the maximum amount of data that may arrive within its time slot, fundamentally avoiding queuing congestion and the resulting periodic delay jitter caused by instantaneous traffic exceeding time slot capacity. Furthermore, through network-level phase orchestration, the message transmission times of each IED are reasonably distributed along the time axis, reducing the probability of multiple periodic data streams arriving at the same port simultaneously, making the traffic smoother in the time dimension and further reducing the amplitude of delay jitter. Based on this, by performing delay measurement and adaptive adjustment during operation, the system can perceive and compensate for the impact of differences in equipment characteristics, clock jitter, or unforeseen disturbances in real time, dynamically optimizing gating parameters to continuously suppress message transmission delay jitter within a preset upper threshold. Ultimately, this can improve the determinism and reliability of critical periodic data stream transmission in the process layer of smart substations, providing a stable and low-jitter communication foundation for high-precision synchronous sampling, phasor calculation, and fast and reliable relay protection. Attached Figure Description

[0021] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0022] Figure 1 This is a schematic diagram illustrating the application of a physical interaction data processing method for power grid equipment according to an example embodiment of this application;

[0023] Figure 2 This is a schematic diagram illustrating the architecture of a physical interaction data processing method for power grid equipment according to an example embodiment of this application;

[0024] Figure 3 This is a flowchart illustrating a method for processing physical interaction data of power grid equipment according to an example embodiment of this application;

[0025] Figure 4 This is a flowchart illustrating an implementation of S200 according to an example embodiment of this application;

[0026] Figure 5 This is a flowchart illustrating an implementation of S200 according to another example embodiment of this application;

[0027] Figure 6 This is a flowchart illustrating an implementation of S200 according to another example embodiment of this application;

[0028] Figure 7 This is a flowchart illustrating an implementation of S400 according to an example embodiment of this application;

[0029] Figure 8 This is a flowchart illustrating an implementation of S400 according to another example embodiment of this application;

[0030] Figure 9 This is a flowchart illustrating an implementation of S400 according to yet another example embodiment of this application;

[0031] Figure 10 This is a flowchart illustrating an implementation of S600 according to an example embodiment of this application;

[0032] Figure 11 This is a flowchart illustrating an implementation of S600 according to another example embodiment of this application;

[0033] Figure 12 This is a flowchart illustrating an implementation of S600 according to yet another example embodiment of this application;

[0034] Figure 13 This is a schematic diagram of the structure of a power grid equipment physical interaction data processing system according to an example embodiment of this application.

[0035] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0036] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0037] Figure 1 This is a schematic diagram illustrating the application of a physical interaction data processing method for power grid equipment according to an example embodiment of this application. For example... Figure 1 As shown in the embodiments of this application, the method provided is applied to the process layer network of newly built or renovated smart substations with voltage levels of 110kV and above. The process layer adopts an industrial Ethernet switch that conforms to the IEC 61850 standard and supports Time-Sensitive Networking (TSN) function, carrying MU sampled value SV messages, GOOSE messages and a small amount of engineering maintenance information.

[0038] In this scenario, the substation is equipped with several sets of intelligent electronic devices (IEDs) such as line protection, transformer differential protection, bus protection, and measurement and control devices. Each IED is connected to the process layer TSN switch via optical fiber or twisted pair. The merging unit periodically generates SV messages at a frequency of 4kHz or higher, and the protection and measurement and control IEDs periodically generate GOOSE status messages. The end-to-end transmission delay jitter is required to be strictly limited to within tens of microseconds to ensure the consistency of sampling reference and the selectivity of action for applications such as differential protection, line longitudinal protection, and synchronous phasor measurement.

[0039] In traditional engineering, gating schedules are typically configured based on experience or average bandwidth, which cannot cover the worst-case scenario where multiple IED message cycles coincide. This leads to SV / GOOSE messages accumulating in the queue on the process layer link during fault disturbances or short-term load fluctuations. The timestamps of the sampled messages observed by the protection device will fluctuate periodically, which in turn causes fluctuations in the protection action time limit or even the risk of false or non-operation.

[0040] This application embodiment obtains the message transmission period and message length of each IED in the station, constructs the flow description information of each SV stream and GOOSE stream, and based on the flow description information, performs worst-case arrival analysis for each output port in the process layer TSN network to determine the maximum amount of data that may need to be sent in each gated time slot under all possible message phase combinations.

[0041] Based on the analysis results, a gating schedule for each port is generated or adjusted to ensure that the available transmission time for each gating slot is not less than the transmission time for the corresponding maximum data volume, thus preventing the formation of queues that cannot be emptied within the gating slots even under extreme alignment conditions. Furthermore, according to the generated gating schedule, this invention assigns specific transmission phase offset parameters to each IED within the station, enabling network-level phase orchestration of the originally independently configured sampling values ​​and GOOSE periodic transmission times across the entire station, reducing the probability of multiple high-priority flows clustering at the same gating slot boundary.

[0042] After the substation is put into operation, during the operation of the TSN network, the transmission delay of various periodic data streams is continuously measured through the local timestamp of the switch, the end-to-end loopback message, or the reception record on the protection device side. This obtains the delay statistics of different IEDs and different service types of messages under long-term operation. When the delay peak or delay jitter in certain time windows is detected to be close to the upper limit of the configuration requirements, the system adaptively adjusts the gating schedule according to the statistical information. For example, it fine-tunes the opening / closing time of certain gating time slots or re-divides the time slot allocation of some data streams, thereby keeping the message transmission delay jitter stably within the preset upper limit threshold.

[0043] By employing the method of this invention in the process layer network of the aforementioned smart substation, key periodic services such as SV / GOOSE can obtain deterministic transmission services even in the worst case without significantly increasing link redundancy and bandwidth configuration. This reduces delay jitter caused by instantaneous congestion at the process layer and improves the time consistency and operational reliability of power secondary systems such as relay protection, measurement and control, and synchronous measurement.

[0044] Figure 2 This is a schematic diagram illustrating the architecture of a physical interaction data processing method for power grid equipment according to an example embodiment of this application. Figure 3 This is a schematic flowchart illustrating a method for processing physical interaction data of power grid equipment according to an example embodiment of this application. Figure 2-3 As shown, the method provided in this embodiment includes:

[0045] S100. Based on the message sending period and message length of each intelligent electronic device (IED) in the process layer, construct the flow description information of each periodic data stream.

[0046] Specifically, the identification information, message sending period, single frame message length, destination MAC address, and VLAN identifier of each periodic data stream can be obtained from the IED's local configuration file, the station control layer engineering database, or the messages in the online subscription handshake process.

[0047] Packets from the same IED that are on the same priority queue, target the same service object, and have the same message sending period are grouped into a periodic data stream. For each periodic data stream, its data stream identifier, message sending period, single frame message length, destination address information, and output port information are recorded.

[0048] Based on the message transmission period and the length of a single frame, the average data rate of the periodic data stream per unit time and the peak data volume within a transmission period are calculated, and the average data rate and peak data volume are stored as part of the flow description information of the periodic data stream.

[0049] Optionally, when constructing the flow description information for each periodic data stream, the message transmission period includes the transmission period of the merge unit sample value SV message and the transmission period of the substation event GOOSE message for general objects.

[0050] S200: Based on the flow description information, perform worst-case arrival analysis for each output port in the TSN.

[0051] In this step, worst-case arrival analysis can be performed on each output port in the TSN based on the flow description information to determine the maximum amount of data that the output port needs to send in each gated slot of the gating schedule under any combination of message phases.

[0052] In current engineering practices, gating schedules are typically configured on switching equipment based solely on average packet bandwidth usage and service priority, with gating time slots statically divided. However, when the phase relationships of multi-source periodic data streams are uncertain, configuration based solely on average bandwidth or empirical margins is insufficient to accurately characterize the instantaneous arrival traffic volume at each output port under the most unfavorable packet phase combination. This results in instantaneous queuing and preemption at the gating time slot level being unpredictable and unavoidable, causing sudden spikes in end-to-end latency of protection and telemetry messages during certain periods, impacting the deterministic communication performance at the process layer.

[0053] In response, Figure 4 This is a flowchart illustrating an implementation of S200 according to an example embodiment of this application. For example... Figure 4 As shown, the above S200 includes:

[0054] S211. Discretize the period of the gating schedule into multiple minimum time units.

[0055] Based on the global time synchronization accuracy and link rate of the time-sensitive network, a minimum time resolution for gating analysis is pre-set. The minimum time resolution is less than or equal to the minimum message transmission interval of various periodic data streams in the time-sensitive network.

[0056] A complete cycle of the gating timetable corresponding to each output port is divided into several consecutively numbered minimum time units. The gating time slot identifier of each minimum time unit and the available transmission bit rate corresponding to that gating time slot are recorded to form a discretized representation of the gating time axis based on the minimum time units.

[0057] S212. Enumerate all possible transmission times for each periodic data stream within its period.

[0058] For each periodic data stream, based on its message transmission period and the initial transmission phase range recorded in the constructed stream description information, the message transmission time is mapped into an equally spaced time sequence with the message transmission period as the step size within the gating time schedule period.

[0059] The sliding phase method is used to shift the initial transmission phase point by point within a message transmission cycle with the minimum time resolution as the step size, and under each initial transmission phase condition, the theoretical transmission time of all messages of the data stream within the gating time cycle is calculated.

[0060] Each theoretical transmission time is quantized according to the minimum time resolution and mapped to the corresponding minimum time unit index to obtain the set of all possible transmission times of the periodic data stream within the period.

[0061] S213. For each output port, calculate the maximum value of the total data volume of each periodic data stream that may arrive within the multiple minimum time units covered by each gated time slot in the gating schedule, and take this maximum value as the maximum data volume.

[0062] For a single output port, within a complete cycle of the gating schedule, the smallest time unit interval covered by each gating slot is traversed sequentially. For each message transmission phase combination, the number of messages of all periodic data streams mapped to the output port within the smallest time unit interval is counted. The number of messages is then multiplied by the length of a single frame of the corresponding data stream and summed to obtain the total amount of data in the smallest time unit interval under that message transmission phase combination.

[0063] After exhaustively or equivalently covering all message transmission phase combinations, the maximum value in the total data volume is selected as the maximum data volume corresponding to the gated time slot. The maximum data volume is then compared with the gated time slot duration and the link rate to verify whether the available transmission time of the gated time slot is not less than the transmission time of the maximum data volume.

[0064] Through the above steps, the period of the gating schedule can be discretized into multiple minimum time units. On this discrete time axis, based on the message transmission period and initial phase range of each periodic data stream, all possible transmission times of each data stream within the gating period are enumerated, and these theoretical transmission times are time-quantized and mapped to specific minimum time units. Based on this, for each output port and each gating time slot, combined with the minimum time unit interval it covers, the number of arriving messages and message lengths of all data streams under different phase combinations are accumulated and statistically analyzed. By traversing or equivalently covering all message phase combinations, the maximum value of the total data volume within each gating time slot is extracted. This constructs a worst-case arrival upper bound in terms of time granularity and phase space. Comparing this upper bound with the gating time slot duration and link rate forms a formal constraint on the security of the gating schedule, realizing a quantitative mapping from data stream period and phase to the available gating service capacity. This enables the available transmission time and link capacity of the gated time slots to substantially cover all possible message phase overlap scenarios, avoiding hidden transmission congestion and out-of-model queuing under certain unforeseen phase combinations, reducing the probability of process layer port delay spikes, and improving the predictability and time determinism of the end-to-end delay upper bound of critical services such as protection and measurement and control.

[0065] Figure 5 This is a flowchart illustrating an implementation of S200 according to another example embodiment of this application. For example... Figure 5 As shown, the above S200 includes:

[0066] S221. For each periodic data stream, construct its arrival curve.

[0067] In this step, an arrival curve can be constructed for each periodic data stream, where the arrival curve represents the maximum amount of data that the data stream can reach within a window of any given time length.

[0068] Specifically, when constructing the flow description information for each periodic data stream, the message sending period, single frame message length, and possible initial sending phase range within a sending period are recorded for each periodic data stream. Based on this, the maximum observation time length, which is not less than the upper limit of the message sending period, is selected and divided into multiple candidate time windows, with each candidate time window corresponding to a window length.

[0069] For each periodic data stream and each window length, the system traverses within the initial transmission phase range with a preset phase step size. Under each initial transmission phase condition, the number of packets that the periodic data stream may arrive within the window length range is calculated, and the product of the number of packets and the length of a single frame packet is taken as the amount of data arriving at the corresponding window length under the initial transmission phase.

[0070] After exhaustively or equivalently covering the initial transmission phase range, the maximum value is selected from multiple arriving data quantities obtained for the window length as the maximum arriving data quantity of the periodic data stream under the window length, thereby forming a discrete arrival curve representing the maximum arriving data quantity within an arbitrary given time length window under different window lengths.

[0071] S222. For each output port, construct its service curve based on the gating schedule.

[0072] In this step, a service curve can be constructed for each output port's gating schedule, where the service curve represents the minimum amount of data that the gating schedule can provide within a window of any given time length.

[0073] Specifically, it can be to obtain the physical link rate connected to the output port in the time-sensitive network and the gating schedule information configured for the output port. The gating schedule information includes at least the opening time, closing time, and service queue priority associated with each gating time slot.

[0074] Based on this, for several different candidate time window lengths, starting from time zero along the time axis, for each candidate time window length, the effective open time intervals of each gated time slot falling within the time range covered by the window are searched one by one.

[0075] The summation of the durations of each effective open time interval is multiplied by the physical link rate to obtain the theoretical maximum amount of data that the output port can send within the candidate time window.

[0076] During the statistical process, gated time slots that do not provide services to the target periodic data stream, as well as time periods that cannot be used for sending the data stream due to guard time deduction, are not included in the effective open time interval. Thus, discrete service curves representing the minimum amount of data sent by the output port within any given time length window are obtained under different time window lengths.

[0077] S223. Check if there is a window of arbitrary length.

[0078] In this step, it can be checked whether there is a window of arbitrary length such that the sum of the arrival curve data volume of all periodic data streams within the window is greater than the service curve data volume. If so, it is determined that the current gating schedule does not meet the worst-case requirements.

[0079] Specifically, for each output port, a set of candidate time window lengths can be selected that are the same as those used when constructing the arrival curve and service curve. For each candidate time window length, the maximum arrival data volume corresponding to each periodic data stream associated with that output port under that window length can be read, and these maximum arrival data volumes can be summed to obtain the sum of the arrival data volumes of all periodic data streams under that window length.

[0080] Simultaneously, the minimum data transmission volume corresponding to the service curve constructed based on the gating schedule within the specified window length is read. The sum of arriving data volumes is compared with the minimum data transmission volume. If, for a certain candidate time window length, the sum of arriving data volumes is greater than the corresponding minimum data transmission volume, it is determined that the current gating schedule configured for this output port has a time window in which it cannot fully absorb the burst of periodic data stream arrivals in the worst-case scenario, thus concluding that the gating schedule does not meet the worst-case requirement. If, for all candidate time window lengths, the sum of arriving data volumes is not greater than the corresponding minimum data transmission volume, it is determined that the gating schedule for this output port meets the service capacity constraints in the worst-case sense.

[0081] In the above steps, for the arrival curve, the number of arriving packets under different initial phase conditions under any given window length is calculated as a set of discrete data using the sending period of each periodic data stream and the length of a single frame, combined with the possible initial phase range. The maximum value is extracted to form the maximum amount of arriving data under that window length, thus forming the upper bound of the function that characterizes the worst arrival amount within any time window.

[0082] For the service curve, based on the physical link rate of a given output port and the opening and closing times of each gated time slot in the gating schedule, the total duration of all valid open time slots providing services to the target service within the coverage area of ​​any given window length is calculated, and multiplied by the link rate to form the minimum amount of service data to be sent under that window length, thus forming a lower bound of the function that characterizes the minimum service capacity within any time window.

[0083] Subsequently, by iterating through a set of candidate time window lengths covering the target analysis range, the maximum amount of data arriving in all data streams under each time window length is summed and compared with the minimum amount of data served by the service curve under that window length. The criterion that the sum of the arrival curves does not exceed the service curve is used to determine whether the gating schedule still has sufficient service capacity under the worst conditions.

[0084] This curve comparison mechanism, based on time window length parameterization, transforms the complex problem of multi-cycle, multi-phase data stream interaction into a computable and verifiable function upper and lower bound relationship check, enabling formal analysis and constraints on whether the gating schedule will cause queue backlog in the worst case.

[0085] Figure 6 This is a flowchart illustrating an implementation of S200 according to another example embodiment of this application. For example... Figure 6 As shown, the above S200 includes:

[0086] S231. Based on flow description information, simulate the arrival and queuing process of periodic data streams in a simulation environment.

[0087] A simulation model consistent with the time-sensitive network topology is built in the configuration center or on a separate simulation server. The simulation model includes simulation nodes and simulation links consistent with the number of intelligent electronic devices, connection relationships, and port bandwidth in the field. The flow description information is imported into the simulation environment. According to the message sending period, message length, and mapped output port of each periodic data flow, the corresponding simulation message source is driven to generate virtual messages in a periodic manner in the simulation environment. The simulated messages are queued and forwarded according to the gating schedule, thereby forming the arrival and queuing behavior corresponding to the actual time-sensitive network operation process.

[0088] S232. Identify and record the gated time slots and corresponding data streams where message queuing occurs during the simulation process.

[0089] In the simulation environment, a queue monitoring and recording structure is established for each gated time slot of each output port to record the changes in queue length before, during, and at the time slot is opened.

[0090] During the simulation, the transmission queues of each output port are continuously monitored. When it is detected that there are packets in the queue that are stuck in the queue and have not been transmitted even after the gating time slot is closed, or when the cumulative amount of packets to be transmitted in the queue at the start of the gating time slot is greater than the amount of data that the gating time slot can provide under the current gating time slot schedule configuration, the gating time slot is marked as a gating time slot with packet queuing, and the periodic data stream identifier that caused the queuing, the corresponding packet arrival time, the packet data amount, and the opening and closing times of the gating time slot are recorded.

[0091] S233. Based on the identification results, iteratively adjust the gating schedule or the sending parameters of the periodic data stream until message queuing no longer occurs in the simulation.

[0092] After obtaining the gating time slots and corresponding data streams of the queuing messages recorded in the simulation environment, the target output port and target gating time slot that need to be adjusted first are determined according to the preset optimization strategy.

[0093] In cases where the gating schedule needs to be adjusted, the available transmission service capacity of the data stream at the corresponding output port can be improved by increasing the duration of the target gating time slot, adjusting the relative position of the target gating time slot within the gating cycle, or adding additional service time slots for the periodic data stream that causes queuing in adjacent gating cycles.

[0094] In situations where it is necessary to adjust the transmission parameters of periodic data streams, the degree of overlap in arrival of multiple periodic data streams in the same gated time slot can be reduced by changing the message transmission phase of the data stream that causes queuing within the allowable range or by increasing the transmission phase offset for non-critical data streams.

[0095] After completing an adjustment of the gating schedule and / or sending parameters, the simulation process is run again in the simulation environment. If message queuing is still detected, the target output port, target gating time slot and corresponding data stream are updated according to the new simulation record. The above adjustment and simulation steps are repeated until message queuing is no longer detected under the preset simulation duration and load conditions, or the preset iteration limit is reached.

[0096] In the above steps, firstly, a simulation model consistent with the field network is built in the simulation environment by the configuration center or simulation server according to the topology, output port configuration and gating schedule of the actual time-sensitive network. The description information of the periodic data stream constructed according to the message sending period, message length and output port mapping is imported into the simulation model, so that each simulated message source generates virtual messages according to the sending period and phase of the actual periodic data stream and queues them for forwarding on the simulation link, thereby reproducing the message arrival and queuing behavior of each output port in the real network.

[0097] Secondly, during the simulation, a queue monitoring and recording structure is established for each gated time slot of each output port. By recording the queue length and the amount of data to be sent at the beginning, end and end of the gated time slot, it is possible to identify situations where there are still messages stuck when the gated time slot is closed or where the amount of data to be sent at the beginning of the gated time slot exceeds the amount of data that the current gated time slot can provide. The corresponding gated time slot and the specific periodic data stream that caused the queuing are identified as objects that need to be optimized.

[0098] Finally, based on the identified target output ports, target gating time slots, and corresponding data streams, the duration of the gating time slots, the position of the gating time slots in the cycle, or the transmission phase of different data streams in the gating schedule are fine-tuned according to a preset strategy. The simulation is rerun after each round of adjustment to verify the results. Through multi-round iterative convergence, the message queuing at each target output port is no longer observed under the predetermined simulation duration and load conditions. This ensures that the gating service capacity and the maximum arrival volume of the periodic data stream meet the constraints, and achieves calculable, verifiable, and convergent control over the worst-case arrival and queuing risks.

[0099] It is worth noting that, regarding the above... Figure 4 The specific implementation shown employs a combined analysis scheme based on time discretization and phase enumeration. By discretizing the gating time period into the smallest time unit, it enumerates all possible transmission times for each periodic data stream within the period, and calculates the maximum amount of data arriving in the periodic data stream within the time unit covered by each gating slot. This maximum value is taken as the worst-case maximum data volume. Compared to... Figure 5 and Figure 6 The specific implementation method shown is as follows:

[0100] First, by discretizing the gating time period into multiple minimum time units and then enumerating all possible transmission times of the periodic data stream within that period, the arrival scenarios of the periodic data stream within the time range covered by each gating slot can be exhaustively considered. This yields the maximum possible data volume that can arrive at a given granularity, achieving fine-scale time-scale analysis tightly coupled with the actual gating time configuration. Second, this scheme calculates directly on the physical time axis of the set of time units covered by the gating slots. The maximum data volume obtained can be directly compared with the available transmission time of the gating slot after link rate conversion, facilitating precise verification and quota configuration for each gating slot. This is suitable for engineering design of precise real-time services such as SV and GOOSE, which have short message periods and are phase-sensitive. Furthermore, compared to abstract network operators or complex simulations, this method can be implemented in engineering using simple discrete-time loops and combinatorial operations. The implementation path is clear and can be directly embedded into the configuration center through static analysis tools. Thus, it achieves a balance between implementation complexity and computational accuracy, making it particularly suitable for configuration verification of small to medium-sized TSN process layer networks with short cycles and fixed gating cycles.

[0101] In response to the above Figure 5 The specific implementation shown involves constructing an arrival curve representing the maximum amount of data arriving within any time window for each periodic data stream, and a service curve representing the minimum amount of service sent within any time window for the gating schedule. Then, by checking whether the sum of the arrival curves exceeds the service curve across all possible time windows, it is determined whether the gating schedule meets the worst-case requirement. This is compared to... Figure 4 and Figure 6 The specific implementation method shown is as follows:

[0102] First, by constructing an arrival curve for the maximum amount of data arriving within an arbitrary time window for each periodic data stream, and a service curve for the minimum amount of service sent within an arbitrary time window for the gating schedule, the discrete queuing problem, which is strongly correlated with the phase and periodic details of specific messages, is transformed into a verification problem of the upper and lower bounds between curves. This achieves a unified formal coverage of all time windows and all phase combinations, ensuring a global analysis of worst-case service arrival. Second, this scheme can inherit the mature analysis tools and proof system of network operator theory. Through the analytical form of curves and affine upper bounds, it performs composable performance analysis on complex cascaded link topologies, thereby supporting the calculation of latency and buffering requirements on end-to-end paths. This is more suitable for the overall performance design of complex multi-hop process layer networks than simply enumerating time slices within a single port. Furthermore, since the analysis results use the curve relationship and the absence of a certain window that makes arrivals greater than service as criteria, the gating configuration margin can be obtained by solving the curve intersection or the maximum difference. This provides a quantifiable analytical safety boundary for subsequent gating time slot configuration and queue capacity setting, which has significant advantages in high-reliability scenarios that require formal verification and proof, such as important hub stations and demonstration projects.

[0103] In response to the above Figure 6 The specific implementation shown involves constructing a virtual network based on flow description information in a simulation environment. Time-step simulations are performed on the arrival and queuing process of periodic data streams to identify the actual gating time slots where queuing occurs and their corresponding data streams. Then, the gating schedule or transmission parameters are iteratively adjusted for these objects until packet queuing no longer occurs during the simulation. This is compared to... Figure 4 and Figure 5 The specific implementation method shown is as follows:

[0104] First, by utilizing a simulation network built based on flow description information and the actual gating schedule, the dynamic evolution of queues within different gating slots of each output port can be directly reproduced in the time domain. This not only detects whether message queuing occurs but also precisely locates the gating slot where queuing occurs and the specific data stream that triggered the queuing, achieving fine-grained fault location and root cause attribution. Second, by combining the simulation process with an iterative adjustment strategy for the gating schedule or transmission parameters, the duration, time position, or data stream transmission phase of the target gating slot are fine-tuned based on the queuing results in each simulation, and this is repeatedly verified until queuing no longer occurs in the simulation. This achieves a simulation-driven closed-loop optimization mechanism, which is closer to actual operating conditions than pure theoretical analysis for scenarios with complex message cycles, non-strictly periodic services, or incomplete field parameters. Furthermore, this solution does not rely on a strict analytical model, making it easy to incorporate engineering factors such as actual equipment processing latency, non-ideal clock deviations, and software stack jitter. These non-ideal factors can be pre-checked during the configuration phase, thereby significantly improving the engineering feasibility and operational robustness of the process layer network after final deployment. It is particularly suitable for performance verification and optimization in large-scale station network transformation or multi-vendor equipment hybrid environments.

[0105] S300: Based on the worst-case arrival analysis results, generate or adjust the gating schedule for each output port in the TSN.

[0106] In this step, based on the results of the worst-case arrival analysis, the gating schedule for each output port in the TSN can be generated or adjusted so that the available transmission time for each gating slot is not less than the transmission time for its corresponding maximum data volume.

[0107] Optionally, the above-mentioned gating schedule for each output port in the TSN includes: deducting the guard time used to prevent frames from crossing the slot boundary when calculating the available transmission time of the gating slot.

[0108] Specifically, for each output port in a time-sensitive network, the maximum message length corresponding to each periodic data stream carried by that output port and the link rate of that output port can be obtained, and the first guard time corresponding to that output port can be determined based on the maximum message length and the link rate.

[0109] When determining the first guard time, the second guard time is also determined based on the message processing delay, clock synchronization error and timestamp quantization error inside the time-sensitive network device. The first guard time and the second guard time are added together to obtain the guard time corresponding to the output port.

[0110] Guard time is reserved near the opening and closing times of each gated time slot, so that no new packets are allowed to be sent in the gated time slot after the guard time begins, and packets that have started to be sent before the guard time arrives can be sent in the current gated time slot. For each gated time slot of each output port in the time-sensitive network, the corresponding guard time is deducted from the duration of the gated time slot, and the remaining time after deduction is used as the available transmission time of the gated time slot.

[0111] During the operation of the time-sensitive network, statistical information on the time interval between the message transmission completion time of the output port and the gating time slot boundary is collected periodically. If the message transmission completion time is detected to be close to the preset threshold of the gating time slot boundary, the guard time of the output port is increased accordingly, and the available transmission time of the gating time slot is recalculated based on the increased guard time.

[0112] S400: Based on the generated gating schedule, assign transmission phase offset parameters to each IED.

[0113] In this step, a transmission phase offset parameter can be assigned to each IED based on the generated gating schedule to perform network-level phase orchestration of the transmission times of the periodic data stream.

[0114] It is worth noting that in current engineering practices, the specific transmission start phase of each IED often adopts the manufacturer's default configuration or is manually set by commissioning personnel based on experience, lacking a unified network-level coordination mechanism. This leads to the possibility that multiple IEDs may simultaneously start transmitting messages at similar times within the gated time slot, causing instantaneous queue accumulation and local congestion. Especially in scenarios with dual-network redundancy or parallel operation of primary and backup channels, the lack of systematic design of the message phase relationship between different channels makes it easy to induce new sudden surge traffic under conditions such as channel switching and fault switching, increasing the risk of process layer network latency jitter and congestion, and affecting the stable acquisition of real-time data by protection and control devices.

[0115] Therefore, given that time slot scheduling has already been implemented at the TSN process layer using a gating schedule, it is necessary to further address how to uniformly orchestrate and dynamically coordinate the transmission phases of periodic messages from each IED at the network level. This would allow the transmission start times of different IEDs within the same gating time slot to be staggered as much as possible, and to establish a controllable fixed phase misalignment relationship between the primary and backup communication channels. This would satisfy the real-time requirements of IEC 61850 services while suppressing instantaneous traffic accumulation within the gating time slots, reducing queuing probability and link congestion risks, and ensuring the consistency and predictability of process layer message timing before and after primary / backup channel switching.

[0116] In response, Figure 7This is a flowchart illustrating an implementation of S400 according to an example embodiment of this application. For example... Figure 7 As shown, the above S400 includes:

[0117] S411. Deploy a scheduling controller at the station control layer.

[0118] The dispatch control software module is pre-installed on the station control layer server of the smart substation, and the server is interconnected with the configuration communication ports of each TSN switch and each IED in the process layer via Ethernet. The dispatch control software module is configured with a time synchronization client connected to the station time synchronization system, so that the dispatch controller can obtain a global time reference consistent with the TSN network devices. The dispatch control software module is also configured with read and write permissions for the message transmission parameters of each IED and the gating schedule of each TSN network device, so as to centrally obtain and distribute configuration information related to the transmission phase offset parameters.

[0119] S412. The scheduling controller runs the arrangement algorithm to calculate and allocate non-conflicting transmission phase offset parameters for each IED.

[0120] The scheduling controller constructs flow description information for each periodic data stream based on the message transmission period, message length, output port identifier, and the opening and closing times of each gated time slot in the corresponding gating schedule collected from each IED.

[0121] Within each gated time slot of the gating schedule, the available transmission time of the gated time slot is divided into multiple non-overlapping transmission phase intervals, based on the transmission period of each periodic data stream and the required transmission time.

[0122] An arrangement algorithm is used to select the start time of the transmission phase interval for each periodic data stream in the gating time table of its output port as its transmission phase offset parameter. This ensures that within any gating time slot, the transmission time intervals corresponding to the transmission phase offset parameters of each periodic data stream mapped to that gating time slot do not overlap on the time axis, thus forming a non-conflicting transmission phase offset parameter allocation result.

[0123] S413, The scheduling controller will send the phase offset parameters to the corresponding IED.

[0124] After calculating the transmission phase offset parameters for each IED, the dispatch controller encapsulates the transmission phase offset parameters corresponding to each IED into a data packet conforming to the preset configuration protocol, and sends it to the management interface of the corresponding IED through the configuration channel between the station control layer and the process layer.

[0125] After receiving a configuration message containing its transmission phase offset parameters, each IED parses the configuration message, writes the parsed transmission phase offset parameters into its local message transmission scheduling configuration table, and updates the transmission start time of the next message transmission cycle based on the global time synchronization result of the time-sensitive network, so that the subsequent periodic data streams are transmitted in the corresponding gated time slots according to the transmission phase offset parameters.

[0126] In the above steps, the station control layer scheduling controller is used as a global coordination node. Through a unified time reference aligned with the TSN time synchronization system, the message transmission period, message length, and gating schedule configuration of the output ports through which each IED passes are obtained. Based on the gating period of the gating schedule and the available transmission time of each gating time slot, the transmission requirements of each periodic data stream on each output port are mapped to the time axis of the gating schedule of that output port. Several non-overlapping transmission phase intervals are divided within each gating time slot, and the arrangement algorithm is run to select a non-overlapping transmission phase offset for each periodic data stream within these intervals.

[0127] Subsequently, the scheduling controller sends the corresponding transmission phase offset parameters of each IED to the local IED through the configuration channel. After aligning with the TSN global clock, the IED determines the start time of periodic message transmission by adding the transmission phase offset to the gating time period. This establishes a one-to-one correspondence between the gating time period, output port, and data stream transmission phase at the protocol level, realizing joint scheduling at the gating granularity and phase granularity. This ensures that within any gating time slot, the transmission time intervals of each periodic data stream are separated on the time axis, thereby suppressing short-term traffic bursts and queue impacts within the time slot.

[0128] Figure 8 This is a flowchart illustrating an implementation of S400 according to another example embodiment of this application. For example... Figure 8 As shown, the above S400 includes:

[0129] S421. Each IED sends messages based on the default transmission phase parameters.

[0130] In the configuration center or station control layer, a default transmission phase parameter corresponding to its message transmission cycle is pre-configured for each IED. After the IED completes time synchronization, based on the default transmission phase parameter and the local clock, the periodic messages are framed at the start of each message transmission cycle and submitted to the TSN network device it accesses, so that the periodic data stream is continuously transmitted according to the default transmission phase parameter when no collision feedback information is received.

[0131] S422. Each IED receives and processes conflict feedback information from network devices.

[0132] In each TSN network device, the queuing delay and queue length of the data stream corresponding to the IED in each output port queue are monitored according to the preset queuing threshold or delay threshold.

[0133] When a queuing delay or queue length exceeding a threshold is detected, a conflict feedback message containing the port identifier of the congestion, the occurrence time, and the identifier of the relevant periodic data stream is generated locally on the TSN network device, and sent to the corresponding IED via management messages, control messages, or maintenance channels. After receiving the conflict feedback message, the IED parses it to determine whether the current transmission phase needs to be adjusted and the range of adjustment required.

[0134] S423. If the conflict feedback information indicates that the current transmission phase is causing congestion, the IED will autonomously adjust its transmission phase offset parameter within the allowable range.

[0135] Based on the congestion occurrence time window and the corresponding gated time slot indicated in the conflict feedback information, the IED determines the position range of the current transmission phase within the gated time schedule period. Without changing the message transmission period, within the pre-configured allowable transmission phase adjustment range, it fine-tunes the current transmission phase offset parameter by delaying or advancing it according to a preset phase fine-tuning step size, generating a new transmission phase offset parameter.

[0136] After completing the adjustment of the transmission phase offset parameters, the IED will redetermine the message transmission start time based on the adjusted transmission phase offset parameters from the next message transmission cycle. During subsequent operation, it will continue to monitor whether it still receives collision feedback information indicating congestion. If no collision feedback information is received within several consecutive observation cycles, the current transmission phase offset parameters will remain unchanged.

[0137] In the above steps, firstly, by pre-configuring a default transmission phase parameter that matches the message transmission period for each IED in the configuration center or station control layer, and driving periodic message framing and transmission according to the default parameter after the IED completes synchronization, it is ensured that each data stream has a definite time position in the gated time slot in the initial state.

[0138] Secondly, by continuously monitoring the queuing delay and queue length of the data stream corresponding to each IED in the queue of each output port in the TSN switching device, once the indicator exceeds the standard within a certain time window, a conflict feedback information containing the congested port identifier, the congestion occurrence time and the relevant periodic data stream identifier is generated locally, and accurately sent to the corresponding IED through management messages or control channels.

[0139] Next, the IED parses the conflict feedback information to locate its current transmission phase position in the gating time period and its relationship with the congestion time window. Without changing the message transmission period, within the preset allowable transmission phase adjustment range, the transmission phase offset parameter is appropriately advanced or delayed according to the preset fine-tuning step size, and transmission is started from the next message period with the adjusted phase.

[0140] Finally, by determining whether congestion feedback for the IED is still being received over multiple observation periods, if the congestion feedback disappears, the current phase is maintained; otherwise, iterative fine-tuning continues within permissible limits. Through this mechanism, the queuing and latency observed at the network layer are quantized into control signals for the IED's transmission phase, forming a flexible time-domain optimization outside the gating schedule. This allows the superposition characteristics of periodic data streams on the time axis to be adaptively rearranged according to the operating state, ensuring the controllability of latency jitter and the determinism of process-layer communication.

[0141] Figure 9 This is a flowchart illustrating an implementation of S400 according to another example embodiment of this application. For example... Figure 9 As shown, the above S400 includes:

[0142] S431. Configure different phase offset modes for the primary and backup communication channels in the process layer network.

[0143] In the configuration center, a first phase offset mode is pre-configured for the primary communication channel and a second phase offset mode is configured for the backup communication channel. The first phase offset mode and the second phase offset mode are each composed of a set of phase offset parameters within a gating time period. Each phase offset parameter in the set of phase offset parameters is used to limit the start time of message transmission of the corresponding IED within the gating time period.

[0144] S432, Assign transmission phase offset parameters based on the first phase offset mode to each IED on the primary communication channel.

[0145] Based on the topology of the primary communication channel and the time length of each gated time slot in the gating schedule, the configuration center selects non-conflicting phase offset parameters from the set of phase offset parameters corresponding to the first phase offset mode for each IED on the primary communication channel. The selected phase offset parameters are then distributed to the corresponding IEDs via the station control layer network, enabling the IEDs to initiate message transmission according to the received phase offset parameters based on their local transmission cycle time.

[0146] S433, Assign transmission phase offset parameters based on the second phase offset mode to each IED on the backup communication channel.

[0147] In this step, each IED on the backup communication channel is assigned a transmission phase offset parameter based on the second phase offset mode, wherein there is a preset fixed misalignment offset between the second phase offset mode and the first phase offset mode.

[0148] Specifically, in the configuration center, based on a fixed offset, the phase offset parameters corresponding to each IED on the primary communication channel are shifted as a whole to form the phase offset parameter set for the backup communication channel. These shifted phase offset parameters are then assigned to each IED on the backup communication channel. This ensures that during a primary / backup switchover between the primary and backup communication channels, each IED on the backup communication channel sends messages within its corresponding gated time slot according to the phase offset parameters shifted from those of the primary communication channel. This reduces the probability of simultaneous overlap of periodic data streams in the primary and backup communication channels when the primary communication channel experiences congestion or failure.

[0149] In the above steps, firstly, a first phase offset mode is established for the primary communication channel in the configuration center, and a second phase offset mode with a fixed offset is established for the backup communication channel. The start time of transmission of each periodic data stream is abstracted as a set of phase offset parameters within the gating time period, and phase allocation is completed for the primary and backup channels respectively. This ensures that each IED in the primary channel is arranged without conflict within its respective gating time slot, while the phase set of the backup channel is a complete translation of the phase set of the primary channel on the time axis.

[0150] Secondly, by setting a fixed offset, the transmission times of the same IED in the primary and backup channels are always separated by a stable time offset within any gating time period. This forms a deterministic time-misaligned redundant scheduling structure during network evolution and multi-cycle operation. Its macroscopic effect is similar to distributing redundant copies of the same service flow in different gating time slots in the time domain.

[0151] Furthermore, from the perspective of queuing theory and worst-case arrival analysis, this staggered phase mode essentially alters the superposition relationship of the arrival functions of redundant data streams on the primary and backup channels. It transforms the originally highly correlated synchronous arrival process into a quasi-independent process with a fixed time offset, ensuring that the composite arrival curve of the output port within any time window is lower than the peak envelope under the original synchronous condition. This formally enhances the margin between the gating service curve and the composite arrival curve, guaranteeing that when the primary channel experiences a momentary traffic peak or jitter amplification, the backup channel will not simultaneously reach its service limit within the same time window, achieving temporal decoupling of the worst-case queuing risk for the primary and backup channels.

[0152] S500 performs delay measurements on periodic data streams during TSN network operation to obtain delay statistics.

[0153] Specifically, at each switching node in the process layer of the smart substation, timestamp marking and recording functions are configured for the output ports carrying periodic data streams. This ensures that when a periodic data stream message arrives at the input queue of the output port, the arrival time information of the message is recorded in the message management information.

[0154] When a message is sent from the output port, the message's sending time information is recorded in the message's management information. Based on the difference between the sending time information and the arrival time information, the queuing and sending delay experienced by the message at the output port is calculated. Within a preset observation period, the delay calculation results of multiple messages belonging to the same periodic data stream are summarized to form the delay statistics of the periodic data stream at the output port.

[0155] It is worth noting that when a link or device failure triggers route reconfiguration or redundancy protection switching during actual operation, the actual forwarding path of the same periodic data stream changes. The worst-case arrival analysis, gating slot available transmission time calculation, and IED transmission phase offset allocation originally performed for the primary forwarding path are only effective for the relevant output ports and queues on the primary path. For the output ports on the backup forwarding path, there is neither a second gating schedule that has been constrained and verified, nor a second transmission phase offset parameter that is aligned with the backup path gating schedule in the time domain.

[0156] If the gating table and transmission phase designed only for the primary path are still used after the failover, it will lead to queue contention, time slot misalignment, and cross-time slot transmission at the backup path nodes. This will cause the queuing delay and jitter of the periodic data stream on the backup path to get out of control, making it impossible to maintain the original deterministic transmission performance.

[0157] To address this, in the configuration center, based on the current network topology and redundancy configuration at the process layer, a primary forwarding path and at least one backup forwarding path can be pre-determined for each periodic data flow with a primary communication channel and a backup communication channel, and the primary forwarding path and at least one backup forwarding path can be associated and stored with the corresponding flow description information.

[0158] For each output port in the TSN, worst-case arrival analysis is performed based on the primary forwarding path and at least one backup forwarding path associated with that output port. A first gating schedule corresponding to the primary forwarding path and a second gating schedule corresponding to each backup forwarding path are generated or adjusted so that, under any forwarding path, the available transmission time of each gating slot is not less than the transmission time of its corresponding maximum data volume.

[0159] Furthermore, based on the first gating schedule and the second gating schedule, a first transmission phase offset parameter corresponding to the primary forwarding path and a second transmission phase offset parameter corresponding to each backup forwarding path are assigned to the corresponding source-end intelligent electronic devices (IEDs). The first and second transmission phase offset parameters are then sent to the IEDs to ensure that the periodic data streams on both the primary and backup forwarding paths are aligned in the time domain with their respective gating schedules.

[0160] During TSN network operation, when a link failure or TSN network device failure related to the primary forwarding path of a periodic data stream is detected and triggers a switchover of the periodic data stream to the corresponding backup forwarding path, the local TSN network device selects and activates the second gating schedule corresponding to the backup forwarding path. It then controls the source-end intelligent electronic device (IED) to switch the transmission phase offset parameter of the periodic data stream from the first transmission phase offset parameter to the corresponding second transmission phase offset parameter, thereby maintaining the deterministic message transmission latency performance of the periodic data stream on the backup forwarding path for a preset time after the fault switchover.

[0161] The above steps, based on the current network topology and redundancy configuration at the process layer, pre-determine a primary forwarding path and at least one backup forwarding path for each periodic data stream with a primary and backup communication channel in the configuration center, and associate and store each forwarding path with its corresponding flow description information. Furthermore, for each output port in the TSN, worst-case arrival analysis is performed based on the primary and backup forwarding paths associated with that output port, generating or adjusting a first gating schedule corresponding to the primary forwarding path and a second gating schedule corresponding to each backup forwarding path. This ensures that the available transmission time for each gating slot under any forwarding path is not less than the transmission time for the corresponding maximum data volume.

[0162] Furthermore, the source-end IED is assigned and distributed a first transmission phase offset parameter corresponding to the primary forwarding path and a second transmission phase offset parameter corresponding to each backup forwarding path, so that the periodic data stream is aligned with its respective gating schedule in the time domain on both the primary and backup forwarding paths. When a fault related to the primary forwarding path is detected during operation and a switchover is triggered, the second gating schedule of the backup path is selected and activated in the local TSN network device, and the source-end IED is controlled to switch the transmission phase offset from the first parameter to the corresponding second parameter.

[0163] Through the aforementioned collaborative mechanism, deterministic message transmission latency performance of periodic data streams can be maintained on the backup forwarding path within a preset time after fault switching. Specifically, this ensures that the maximum end-to-end latency and jitter indicators of such critical services are continuously controlled before and after fault switching, avoiding sudden increases in latency and uncontrolled jitter during the switching window, thereby solving the problem that the backup channel cannot provide deterministic transmission.

[0164] S600 adaptively adjusts the gating schedule based on delay statistics.

[0165] In this step, the gating schedule is adaptively adjusted based on latency statistics to keep message transmission latency jitter within a preset upper threshold.

[0166] It is worth noting that in actual operation, due to factors such as changes in the operating status of primary equipment within the station, changes in the frequency of protection logic actions, firmware upgrades of some IEDs, or adjustments to the plant-station communication methods, the actual arrival pattern of periodic services at the process layer will deviate from the ideal period and fixed phase model assumed during design. Messages are more likely to cluster and queue near the gated time slot boundary, resulting in periodic spikes and non-negligible delay jitter in end-to-end transmission delay.

[0167] Existing solutions mostly only set the gating schedule and priority queue once during the design phase, and lack a closed-loop adjustment mechanism based on operational delay statistics. This makes it impossible to specifically correct and suppress the time slot boundary effect, queue congestion mode and abnormal jitter of specific service flows that gradually appear under long-term network operation. As a result, it brings potential risks of maloperation and failure to operate to secondary systems such as relay protection, measurement and control and synchronous measurement that are highly sensitive to time determinism.

[0168] In response, Figure 10 This is a flowchart illustrating an implementation of S600 according to an example embodiment of this application. For example... Figure 10 As shown, the above-mentioned S600 includes:

[0169] S611. Periodically or during preset low-load periods, report delay statistics to the configuration center.

[0170] In each TSN network device, a reporting period and low load judgment conditions are pre-configured. When it is detected that the current time meets the reporting period, or the current network link utilization, CPU utilization and packet queue length are all lower than the preset low load threshold, the latency measurement results stored locally according to the output port and periodic data stream are read. Within the preset observation time window, the maximum latency, minimum latency, average latency and latency jitter and other latency statistics are calculated, encapsulated into a latency statistics reporting message, and securely transmitted to the configuration center through the management channel or operation and maintenance network.

[0171] S612. The configuration center analyzes the latency statistics. If periodic latency spikes are identified, the adjusted gating schedule is recalculated and generated.

[0172] After receiving latency statistics reports from multiple TSN network devices, the configuration center organizes and archives the latency statistics information according to the device identifier, output port identifier, and periodic data stream identifier. Within a preset analysis period, time-series analysis is performed on the latency statistics of each output port. Based on the changes in maximum latency and latency jitter within multiple consecutive observation windows, latency spikes that recur within fixed time intervals are identified. If the latency spike is determined to have periodic characteristics related to the gating schedule period, the gating schedule calculation module is invoked. Based on the original gating schedule and the description information of the current service flow, the opening and closing times of the gating time slots and the time slot lengths of the output ports that generated the latency spikes are reallocated, resulting in an adjusted gating schedule. The legality and consistency are then verified locally.

[0173] S613, the configuration center will send the adjusted gating schedule to the corresponding TSN network devices.

[0174] In the configuration center, the verified and adjusted gating schedule is associated with the corresponding TSN network devices and output ports to generate a configuration distribution message containing information such as the gating schedule identifier, application effective time, and rollback policy. This message is sent to the corresponding TSN network devices via a pre-established configuration management channel. Upon receiving the message, the TSN network devices parse and verify the gating schedule configuration information. After the current message sending cycle ends, gating scheduling is switched to the adjusted gating schedule according to the application effective time. New latency statistics are collected and reported during a preset observation period to allow the configuration center to verify the effectiveness of the adjusted gating schedule in suppressing periodic latency spikes.

[0175] In the above steps, firstly, on the side of each TSN network device, by recording timestamps in the packet forwarding path or using a hardware timestamp mechanism, the queuing delay of packets belonging to the same periodic data stream at the egress port is measured, and within a preset observation window, statistics such as maximum delay, average delay, and delay jitter are calculated and stored according to the output port identifier and data stream identifier.

[0176] Secondly, on the configuration center side, it periodically receives latency statistics reporting messages from multiple TSN devices, constructs a latency statistics sequence of each output port over time, uses algorithms such as window sliding analysis and peak detection to identify latency spikes with fixed periodic characteristics, and compares this period with the current gating time schedule period to determine whether the spike is caused by a mismatch between the gating time slot configuration and the arrival characteristics of the service flow.

[0177] Secondly, after confirming that the delay spike is related to the gating schedule, the configuration center calls the gating schedule calculation module. Based on the current flow description information and the detected spike time position, it adjusts the gating time slot opening time, closing time and duration of the relevant output ports so that the minimum sending capacity under the service curve in any window covers the maximum arrival volume reflected in the delay statistics, and performs legality, conflict and end-to-end link consistency checks.

[0178] Finally, the adjusted gating schedule and effective time policy are distributed to the corresponding TSN devices via the configuration management channel. The devices then switch gating configurations at appropriate cycle boundaries and continue to report new latency statistics during subsequent operation to verify whether latency spikes have been effectively eliminated or significantly reduced, thus achieving iterative optimization and convergence control of the gating schedule. This adaptive gating adjustment mechanism based on latency statistics feedback transforms the gating schedule from a statically determined design to a dynamically convergent one during operation, ensuring in principle that the process layer can maintain calculable, verifiable, and controllable deterministic latency performance under business evolution and changes in operating conditions.

[0179] Figure 11 This is a flowchart illustrating an implementation of S600 according to another example embodiment of this application. For example... Figure 11 As shown, the above-mentioned S600 includes:

[0180] S621. On the local side of the TSN network device, based on the delay statistics information within the preset observation window, determine whether there are delay spikes associated with specific gated time slot boundaries.

[0181] At the output port of the TSN network device, the time when packets belonging to the same periodic data stream enter the port queue and the time when they leave the port queue are recorded. The queuing delay of a single packet transmission is calculated. Within multiple observation windows divided based on the gating time period, the queuing delay is statistically analyzed to obtain the maximum delay, average delay, and delay jitter within each observation window.

[0182] Then, the delay statistics in each observation window are folded and mapped according to the period of the gating time schedule. The delay values ​​mapped to the same time position in the same gating time schedule period are aggregated and analyzed. When the aggregated delay value near the opening or closing time of a certain gating time slot is consistently higher than the preset delay spike threshold, it is determined that there is a delay spike associated with the boundary of the gating time slot.

[0183] S622. If it exists, the opening or closing time of the specific gated time slot is shifted within a preset fine-tuning step range.

[0184] In the locally stored gating schedule, specific gating slots identified as being associated with latency spikes are selected. The opening or closing time of these gating slots is then shifted forward or backward in time with a preset fine-tuning step. The fine-tuning step is less than one-tenth of the gating schedule period, ensuring that the duration of the shifted gating slot is not less than the minimum transmission time required by the corresponding priority queue under worst-case arrival analysis. Simultaneously, before adjustment, conflict checks are performed on the opening and closing times of other gating slots on the same output port to prevent overlap or overlap between gating slots of different priorities on the timeline.

[0185] S623. After translation, continue to observe the delay change. If the delay peak decreases, retain the adjustment; otherwise, restore the original state and try to translate in the opposite direction.

[0186] After shifting the opening or closing times of a specific gated time slot, delay statistics are collected within a preset number of periods using the same observation window length and statistical method as before the adjustment. The shifted delay statistics are then compared with the delay statistics at the same gated time period position before the adjustment. When both the maximum delay and delay jitter after the shift are lower than the preset improvement threshold, the delay spike is determined to be weakened or eliminated, and the current gated time schedule configuration remains unchanged as the new baseline state.

[0187] If the maximum delay or delay jitter does not decrease significantly after the shift or even increases, the gating timetable is restored to its original state before adjustment. While maintaining the same fine-tuning step size, the opening or closing time of the specific gating slot is shifted in the reverse direction, and the above observation and comparison process is repeated until a gating slot boundary position that can reduce delay spikes is found in either the forward or reverse shift. Alternatively, if the preset maximum number of fine-tuning attempts is reached, and the delay spike is not significantly reduced after reaching the maximum number of attempts, the identifier of the output port and the gating slot is recorded, and relevant delay statistics are reported to the configuration center for further analysis.

[0188] In the above steps, firstly, at the output port of the TSN network device, the timestamps of each packet entering the port queue and leaving the port queue are recorded for each packet belonging to the same periodic data stream. The queuing delay of a single packet is calculated, and multiple observation windows are divided based on the gating time period. The maximum delay, average delay and delay jitter in each window are statistically analyzed.

[0189] Secondly, the delay statistics of each observation window are folded and mapped according to the gating time period, that is, the time series spanning multiple periods is compressed into a single period time axis. The delay values ​​mapped to the same time position in the same gating period are aggregated and analyzed to identify the time positions where the delays are clustered near a specific gating boundary and are continuously higher than the preset peak threshold.

[0190] Then, for the gated time slots identified as being associated with latency spikes, the opening or closing time of the gated time slot is slightly shifted in the locally stored gated time schedule. The fine-tuning step size is strictly limited to a small proportion of the gated period. Before and after the fine-tuning, it is checked whether there is any overlap or coverage between the gated time slots to maintain the feasibility of the overall gated structure.

[0191] Finally, after completing the boundary shift, the same observation window and folding mapping method are used to statistically analyze the new delay distribution. The fine-tuned maximum delay and delay jitter are compared with the data before adjustment. If a delay spike is detected to have significantly weakened or disappeared, the adjustment is retained; otherwise, the original configuration is restored and a reverse shift is attempted. Through this local closed-loop mechanism, the gated time slot boundary position can adaptively converge to a more balanced delay distribution during operation, thereby achieving dynamic suppression of periodic delay spikes without changing the service configuration.

[0192] Figure 12 This is a flowchart illustrating an implementation of S600 according to another example embodiment of this application. For example... Figure 12 As shown, the above-mentioned S600 includes:

[0193] S631. Based on latency statistics, identify specific data streams whose latency jitter continuously exceeds a threshold.

[0194] Locally on the TSN network device, a sliding observation window is maintained for each periodic data stream. For each packet belonging to the same periodic data stream, the timestamps of its entry into and exit from the queue are recorded. The queuing delay of a single packet is calculated. Within each observation window, the maximum queuing delay, minimum queuing delay, and average queuing delay of the periodic data stream are calculated respectively. The difference between the maximum queuing delay and the minimum queuing delay is used as the delay jitter metric.

[0195] When the latency jitter metric is greater than the preset latency jitter threshold in multiple consecutive observation windows, the corresponding periodic data stream is marked as a specific data stream whose latency jitter continuously exceeds the threshold.

[0196] S632. Locally on the TSN network device, dynamically map a specific data stream to a higher priority transmission queue.

[0197] Multi-level gated queues or priority queues are pre-configured in TSN network devices, and a corresponding priority identifier is configured for each level of queue.

[0198] When a periodic data stream is marked as a specific data stream, without changing the service identifier and message format of the periodic data stream, the queue selection item of the periodic data stream in the local forwarding table or queue mapping table is updated, and the original mapped sending queue is replaced with a sending queue with a higher priority identifier, so that the messages of the periodic data stream that arrive later will be given priority in the queuing process over the messages in the lower priority queue.

[0199] After completing the queue mapping update, the latency jitter metric of the specific data stream continues to be counted within the sliding observation window. When it is detected that the latency jitter metric is continuously lower than the latency jitter threshold and maintains a preset stable period, the queue mapping can be optionally restored to the original priority queue.

[0200] S633, divert a portion of the traffic of a specific data stream to another physical port for transmission.

[0201] In TSN network devices, flow matching rules containing the target MAC address, VLAN identifier, Ethernet priority and ingress port information are maintained for each periodic data flow, and redundant or backup links that can be used to carry the periodic data flow are pre-configured for the target physical port in the forwarding plane of the TSN network device.

[0202] When a periodic data stream is marked as a specific data stream and its latency jitter continues to exceed a threshold even after increasing the queue priority, a traffic splitting rule is generated for that specific data stream in the forwarding plane. A subset of packets that meet the preset traffic splitting conditions, including but not limited to packets arriving within a specific time slice, packets with odd or even sequence numbers, or packets exceeding the preset instantaneous arrival rate, are mapped to the sending queue corresponding to another physical port. The remaining packets are still sent through the original physical port.

[0203] After the traffic split takes effect, the end-to-end latency and latency jitter of the specific data stream on the two physical ports will be independently statistically analyzed. If the overall latency jitter metric of the specific data stream is found to be continuously lower than the latency jitter threshold under the condition that the split ratio and split conditions remain unchanged, the current split strategy will be maintained. Otherwise, the split ratio and split conditions will be adjusted or another physical port will be selected for splitting based on the statistical results.

[0204] In the above steps, firstly, an independent latency statistics object is established locally on the TSN network device for each periodic data stream. By recording the timestamps of entering and leaving the queue, the maximum queuing latency, minimum queuing latency, and average queuing latency of the data stream within the sliding observation window are continuously calculated. The latency jitter metric is constructed from the difference between the maximum and minimum queuing latency. Data streams that exceed the threshold for multiple consecutive windows are marked as specific data streams, thereby realizing the time series identification of continuous abnormal jitter behavior.

[0205] Secondly, by utilizing the programmable queue mapping capability of the TSN switching chip, without changing the Ethernet header identifier of the data stream, the corresponding sending queue is adjusted only in the local forwarding table or queue mapping table, so that it is migrated from the original queue to a queue with higher priority and stronger service guarantee. The queuing delay fluctuation is directly reduced by using higher priority gating time slots or more abundant service bandwidth.

[0206] Furthermore, when queue escalation still fails to meet jitter constraints, the forwarding capability of the switching chip's output port and the pre-configured redundant links are utilized. By adding time-, sequence-, or arrival rate-based traffic splitting conditions to the forwarding rules, a portion of the packets of this specific data stream is mapped to another physical port, decoupling it from the high-load traffic on the original port on the physical path, thus fundamentally weakening the periodic superposition effect on a single output port.

[0207] In this embodiment, flow description information for each periodic data stream is constructed based on the message transmission period and message length of each intelligent electronic device (IED) in the process layer. Based on the flow description information, worst-case arrival analysis is performed for each output port in the TSN. Based on the results of the worst-case arrival analysis, a gating schedule for each output port in the TSN is generated or adjusted. Based on the generated gating schedule, a transmission phase offset parameter is assigned to each IED. During the operation of the TSN network, the delay of the periodic data stream is measured to obtain delay statistics. Based on the delay statistics, the gating schedule is adaptively adjusted to keep the message transmission delay jitter within a preset upper limit threshold, thereby improving the determinism and reliability of critical periodic data stream transmission in the process layer of the intelligent substation.

[0208] Figure 13 This is a schematic diagram illustrating the structure of a power grid equipment physical interaction data processing system according to an example embodiment of this application. For example... Figure 13 As shown, the power grid equipment physical interaction data processing system 700 provided in this embodiment includes:

[0209] Module 710 is used to construct flow description information for each periodic data stream based on the message transmission period and message length of each intelligent electronic device (IED) in the process layer of a smart substation using Time-Sensitive Networking (TSN).

[0210] Execution module 720 is used to perform worst-case arrival analysis for each output port in the TSN based on the flow description information;

[0211] The generation module 730 is used to generate or adjust the gating schedule for each output port in the TSN based on the results of the worst-case arrival analysis.

[0212] Transmission module 740 is configured to assign transmission phase offset parameters to each of the IEDs based on the generated gating schedule;

[0213] The measurement module 750 is used to perform delay measurement on the periodic data stream during the operation of the TSN network to obtain delay statistics.

[0214] The adjustment module 760 is used to adaptively adjust the gating schedule based on the delay statistics.

[0215] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this application are indicated by the claims.

[0216] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.

Claims

1. A method for processing physical interaction data of power grid equipment, characterized in that, The method is applied to the process layer of a smart substation employing Time-Sensitive Networking (TSN); the method includes: Based on the message transmission period and message length of each intelligent electronic device (IED) in the process layer, the flow description information of each periodic data stream is constructed. Based on the flow description information, worst-case arrival analysis is performed for each output port in the TSN; Based on the results of the worst-case arrival analysis, generate or adjust the gating schedule for each output port in the TSN; Based on the generated gating schedule, a transmission phase offset parameter is assigned to each IED to perform network-level phase orchestration of the transmission times of the periodic data stream; During the operation of the TSN network, delay is measured on the periodic data stream to obtain delay statistics. Based on the aforementioned delay statistics, the gating schedule is adaptively adjusted.

2. The method according to claim 1, characterized in that, Based on the flow description information, the worst-case arrival analysis is performed for each output port in the TSN, including: The period of the gated schedule is discretized into multiple minimum time units; Enumerate all possible transmission times for each periodic data stream within the period; For each output port, calculate the maximum value of the total data volume of each periodic data stream that may arrive within the plurality of minimum time units covered by each gated time slot in the gated schedule, and use this maximum value as the maximum data volume.

3. The method according to claim 1, characterized in that, Based on the flow description information, the worst-case arrival analysis is performed for each output port in the TSN, including: Based on the flow description information, the arrival and queuing process of the periodic data stream is simulated in a simulation environment; Identify and record the gated time slots where message queuing occurs during the simulation process and the corresponding data streams; Based on the identification results, the gating schedule or the sending parameters of the periodic data stream are iteratively adjusted until the message queuing no longer occurs in the simulation.

4. The method according to claim 1, characterized in that, The process of assigning transmission phase offset parameters to each IED based on the generated gating schedule includes: Deploy a scheduling controller at the station control layer; The scheduling controller runs a deployment algorithm to calculate and allocate non-conflicting transmission phase offset parameters for each IED; The scheduling controller sends the transmission phase offset parameter to the corresponding IED.

5. The method according to claim 1, characterized in that, The process of assigning transmission phase offset parameters to each IED based on the generated gating schedule includes: Each of the IEDs transmits messages based on the default transmission phase parameters; Each of the IEDs receives and processes conflict feedback information from network devices; If the conflict feedback information indicates that the current transmission phase is causing congestion, the IED will autonomously adjust its transmission phase offset parameter within an allowable range.

6. The method according to claim 1, characterized in that, The adaptive adjustment of the gating schedule based on the delay statistics includes: The latency statistics are reported to the configuration center periodically or during preset low-load periods. The configuration center analyzes the latency statistics and, if periodic latency spikes are identified, recalculates and generates an adjusted gating schedule. The configuration center will send the adjusted gating schedule to the corresponding TSN network devices.

7. The method according to claim 1, characterized in that, The adaptive adjustment of the gating schedule based on the delay statistics includes: Locally on the TSN network device, based on the latency statistics within a preset observation window, it is determined whether there are latency spikes associated with specific gated time slot boundaries; If it exists, the opening or closing time of the specific gated time slot will be shifted within a preset fine-tuning step range; After translation, continue to observe the delay changes. If the delay peak decreases, retain the adjustment; otherwise, restore the original state and try to translate in the opposite direction.

8. The method according to any one of claims 1-7, characterized in that, When constructing the flow description information for each periodic data stream, the message transmission period includes the transmission period of the merge unit sampled value SV message and the transmission period of the substation event GOOSE message for general objects.

9. The method according to any one of claims 1-7, characterized in that, The generation or adjustment of the gating schedule for each output port in the TSN specifically includes: deducting the guard time used to prevent frames from crossing the slot boundary when calculating the available transmission time of the gating slot.

10. A physical interaction data processing system for power grid equipment, characterized in that, include: The module is used to construct the flow description information of each periodic data stream based on the message transmission period and message length of each intelligent electronic device (IED) in the process layer of a smart substation using Time-Sensitive Networking (TSN). The execution module is used to perform worst-case arrival analysis for each output port in the TSN based on the flow description information; The generation module is used to generate or adjust the gating schedule for each output port in the TSN based on the results of the worst-case arrival analysis. The transmitting module is used to assign a transmitting phase offset parameter to each of the IEDs based on the generated gating schedule; The measurement module is used to measure the delay of the periodic data stream during the operation of the TSN network and obtain delay statistics. The adjustment module is used to adaptively adjust the gating schedule based on the delay statistics.

Images