Intelligent substation communication message transmission method based on time-aware shaping

Abstract

The invention provides an intelligent substation communication message transmission method based on time-aware shaping. The method comprises the following steps: distributing a fixed periodic time slice for a periodic sampling value SV message in an intelligent substation through time aware shaping TAS, and transmitting the SV message in the periodic time slice; the weight value of the queue is dynamically adjusted by monitoring the total byte number of the aperiodic service message to be sent in the queue in real time, and the aperiodic service message in the queue is scheduled and transmitted by adopting a dynamic weighting DW-TAS scheduling algorithm based on time-aware shaping according to the weight value of each queue. According to the method provided by the invention, the more critical sudden GOOSE traffic can be transmitted as much as possible under the condition of burst load on the premise of ensuring the basic transmission requirement of the secondary critical traffic, and meanwhile, the deterministic transmission of the periodic SV traffic can be realized by distributing a fixed transmission time slot for the periodic SV traffic.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent substation message transmission, and in particular to a method for transmitting communication messages in an intelligent substation based on Time-Aware Shaper (TAS). Background Art

[0002] In recent years, with the continuous development of information technology, substations have shown an intelligent development trend. Researchers have proposed a scheme for unified networking of the entire intelligent substation, that is, all intelligent electronic devices (IEDs) in the intelligent substation are connected to the same network. This networking method can achieve data integration to the greatest extent, improve operation efficiency and maintenance costs, but at the same time brings problems such as network security, device compatibility, and data management. In addition, new challenges are also posed to the real-time performance and reliability of communication.

[0003] In an intelligent substation, there are various communication service messages, and various communication services have different traffic characteristics and communication service requirements. These mainly include Generic Object Oriented Substation Event (GOOSE) messages, Sample Value (SV) messages, and Manufacturing Message Specification (MMS) messages. The GOOSE messages include trip commands, switch position change information, and device status information. The SV messages are file acquisition information and belong to periodic data streams. The MMS messages transmit file transfer information, which is mainly used to transmit large file data such as records and settings, and there is no specific delay requirement. When a substation fails, a large number of trip commands and switch position change information will be generated. Therefore, when a failure occurs, the queue scheduling strategy needs to be adjusted to ensure the delay requirements of various communication services.

[0004] Currently, in the actual project of an intelligent substation, two levels of priorities are defaultly adopted. The messages for general substation events and sample value messages are set as high priorities, and the remaining substation communication services are set as low priorities. And the strict priority queue (SPQ) scheduling algorithm is used on the switch. However, this simple priority division method is difficult to meet the real-time requirements of different types of services in the substation.

[0005] At present, there has been some research on the substation service queue scheduling strategy in the prior art. Some solutions use the SPQ algorithm to conduct a more detailed priority division for different information types. Although it can ensure the low-latency transmission of time-sensitive information such as trip commands and SV messages, it increases the possibility of latency accumulation of low-priority messages in case of congestion. There are also solutions that adopt a two-level scheduling strategy, dividing GOOSE messages, sampled value messages, device status information, and file transfer information into 4 priorities from high to low, and using the SPQ algorithm to preferentially schedule the bursty GOOSE messages. The current traditional traffic scheduling methods, although they can meet the priority requirements of different services in the substation to a certain extent, cannot ensure the deterministic scheduling of SV messages and are difficult to cope with the highly dynamic and complex communication environment.

[0006] Time-Sensitive Networking (TSN) is a series of standards promulgated by the Institute of Electrical and Electronics Engineers (IEEE). TSN can achieve the real-time and reliable transmission of data traffic in the Ethernet physical facilities through various traffic scheduling and shaping mechanisms. TSN technology can improve the stability of the power system. In terms of substation traffic scheduling of TSN, current scholars mostly use TAS to control the switch state of the queue gate through the Gate Control List (GCL) to achieve the substation service traffic scheduling. Some solutions have proposed a traffic scheduling method combining TAS and frame preemption. This method reduces the transmission collision on the link and improves the transmission latency of time-sensitive flows in the substation, providing a new idea for the improvement of the Quality of Service (QoS) of the power system. There are also solutions that propose a maximum latency calculation method under the TAS technology based on network calculus, providing theoretical support for the application of the TAS technology in intelligent substations.

[0007] The disadvantages of the TAS scheduling mechanism for messages in the above prior art include: The TAS scheduling mechanism has limitations in coping with the latency requirements of non-periodic data streams. Especially when dealing with bursty GOOSE traffic messages, it is difficult to design a reasonable gate control list for all traffic. If the gate control list is set improperly, it may even lead to more serious latency problems. Therefore, applying the TAS scheduling algorithm alone is difficult to meet the complex requirements of substation communication services. Summary of the Invention

[0008] An embodiment of the present invention provides a method for transmitting communication messages of an intelligent substation based on TAS to effectively improve the transmission efficiency of communication messages of the intelligent substation.

[0009] To achieve the above object, the present invention adopts the following technical solutions.

[0010] An intelligent substation communication message transmission method based on time-aware shaping, comprising:

[0011] Allocating a fixed periodic time slice for the periodic sampled value SV message in the intelligent substation through time-aware shaping TAS, and transmitting the SV message within the periodic time slice;

[0012] Dynamically adjusting the weight value of the queue by real-time monitoring the total number of bytes of the aperiodic service messages to be sent in the queue, and scheduling and transmitting the aperiodic service messages in the queue using a dynamic weighted DW-TAS scheduling algorithm based on time-aware shaping according to the weight values of each queue.

[0013] Preferably, the step of allocating a fixed periodic time slice for the periodic sampled value SV message in the intelligent substation through time-aware shaping TAS and transmitting the SV message within the periodic time slice includes:

[0014] Using a gating list to divide the communication time into a periodic time slice and an aperiodic time slice. After the nth cycle SV message enters the buffer queue, it is transmitted within the periodic time slice of cycle n, and the remaining aperiodic service messages are transmitted within the aperiodic time slice. A guard band is set before the start of the periodic time slice, and the size of the guard band is determined according to the delay required for the Ethernet maximum transmission unit data packet to be transmitted in the switch.

[0015] Preferably, the step of allocating a fixed periodic time slice for the periodic SV message through TAS and transmitting the SV message within the periodic time slice includes:

[0016] Let the length of the SV message be L SV bytes, and there are n messages passing through each cycle. Then the size T of the periodic time slice is:

[0017]

[0018] In the formula: R is the link transmission rate;

[0019] Assume that each application protocol data APDU in the message includes 4 application service data units ASDU, that is, it contains 4 voltage and current sampling data. There are 4 SV messages passing through the switch in each sampling cycle of the merging unit. At this time, the frame length of the SV message is 160 Byte, that is:

[0020]

[0021] At this time, the size of the periodic time slice is:

[0022]

[0023] Preferably, dynamically adjusting the weight value of the queue by monitoring the total number of bytes of the aperiodic service packets to be sent in the queue in real time includes:

[0024] Set Q 1 , Q 2 ,..., Q m represent m queues on the network node, corresponding to m types of traffic in the network respectively. Q 1 caches file transfer information and has non-real-time requirements. Q 2 , Q 3 , Q 4 , Q 5 respectively cache the trip command, switch position change information, device status information and time synchronization packets with real-time requirements;

[0025] Dynamically adjust the weight value of the queue by monitoring the total number of bytes of the packets to be sent in the queue in real time. The specific rules are as follows:

[0026] (1) When the total number of bytes of the packets in the queue is lower than the set lower limit value MinByte i , the queue weight remains unchanged at the initial value;

[0027] (2) When the total number of bytes of the packets is higher than the set upper limit value MaxByte i , the queue weight is adjusted to the preset maximum value to give priority to ensuring bursty traffic;

[0028] (3) When the total number of bytes of the packets is between the upper limit MaxByte i and the lower limit MinByte i , the weight value is dynamically adjusted according to the adjustment formula to make full use of the bandwidth resources and ensure the timely transmission of bursty traffic. The adjustment formula is shown in Equation (5):

[0029]

[0030] where init_wt i is the initial weight value of queue i, max_wt i is the maximum weight of queue i, Byte i is the total number of bytes of the packets to be sent in the queue, MinByte i and MaxByte i are the set upper and lower limit values, and k is the weight adjustment coefficient of queue i;

[0031] Q 4 , Q 5 The weight of the queue is dynamically adjusted according to Equation (5). The weight adjustment rule of Q 3 is shown in Equation (6);

[0032] wt 3 = 1 - wt 4 -wt 5 -wt 2 (6)

[0033] The initial weights of the queues are init_wt 5 = 0.2, init_wt 4 = 0.2, init_wt 3 = 0.5, init_wt 2 = 0.1, and for Q 4 , Q 5 the weight adjustment coefficient k of 4 = 1, k 5 = 3, and the corresponding maximum weights max_wt 5 = 0.5, max_wt 4 = 0.3, for Q 3 the minimum weight of Q is min_wt 3 = 0.1, for Q 2 the weight of Q is fixed at wt 2 = 0.1.

[0034] Preferably, the non-periodic service packets in the queue are scheduled and transmitted by using a dynamic weighted DW-TAS scheduling algorithm based on time-aware shaping according to the weight values of each queue, including:

[0035] The scheduling algorithm for non-periodic packet traffic in the dynamic weighted DW-TAS scheduling algorithm is as shown in Equation (4).

[0036]

[0037] VQ 1 is the virtual queue rearranged from Q 2 ~Q 5 according to the DWRR scheduling rule. VQ is the virtual queue rearranged from VQ according to the SPQ scheduling rule 1 and Q 1 and is the final output queue. The priority of VQ 1 is higher than the priority of Q 1 ;

[0038] Within a polling cycle, the total number of bytes that all queues can schedule is Total byte , then the bandwidth occupancy ratio of each queue is its weight value, satisfying the following relationship:

[0039]

[0040] Queue Q iThe number of bytes that can be sent within a polling period: SendByte i As shown in (8):

[0041] SendByte i = Total byte * wt i (8)

[0042] After the message arrives, it enters the corresponding queue according to its type. During the periodic time slice, the switch transmits the SV message. During the aperiodic time slice, when there are data packets waiting in Q 2 ~Q 5 and there are data packets waiting, the scheduling algorithm uses the polling scheduling mechanism to schedule the data packets in Q 2 ~Q 5 sequentially, and calculates the waiting bytes of the data packets in each queue during each round of scheduling, and dynamically updates the weights according to formulas (5) and (6), and calculates the corresponding number of bytes to be sent: SendByte i , and sends the data packets according to SendByte i . When there are no data packets waiting in queues 2 to 5, the algorithm processes the file transfer information in the last queue according to the SPQ algorithm.

[0043] As can be seen from the technical solutions provided by the embodiments of the present invention described above, the method of the present invention can effectively ensure the deterministic transmission of the periodic SV data stream in the case of dealing with bursty GOOSE messages, while optimizing the end-to-end delay performance of bursty traffic, and can better meet the actual requirements of traffic scheduling for intelligent substation services.

[0044] Additional aspects and advantages of the present invention will be given in part in the following description, which will become apparent from the following description, or can be understood through the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0046] Figure 1 FIG. is a schematic diagram of the division of communication time provided by the embodiment of the present invention;

[0047] Figure 2 FIG. is a schematic diagram of the principle of the TAS-PDWRR queue scheduling algorithm provided by the embodiment of the present invention;

[0048] Figure 3The present invention provides a scheduling algorithm flowchart for an intelligent substation communication message transmission method based on time-aware shaping;

[0049] Figure 4 It is a simulation model structure diagram of a D2-1 type intelligent substation provided for an embodiment of the invention;

[0050] Figure 5 It is a schematic diagram of a star topology structure abstracted from the communication service flow of an intelligent substation provided for an embodiment of the invention;

[0051] Figure 6 It is a histogram of the delay distribution of SV message transmission by each algorithm provided for an embodiment of the invention;

[0052] Figure 7 It is a box plot of the delay jitter of different scheduling algorithms under traffic burst conditions provided for an embodiment of the invention;

[0053] Figures 8 - 10 It respectively shows the average end-to-end delay of several algorithms in the case of GOOSE message bursts to different degrees. Detailed implementation manners

[0054] The following details the implementation manners of the present invention. Examples of the implementation manners are shown in the drawings, where the same or similar reference numerals represent the same or similar elements or elements with the same or similar functions from beginning to end. The implementation manners described below with reference to the drawings are exemplary and are only used to explain the present invention and should not be construed as a limitation of the present invention.

[0055] Those skilled in the art of the present technology can understand that unless specifically stated otherwise, the singular forms "a", "an", "the" and "said" used herein may also include the plural forms. It should be further understood that the term "comprising" used in the specification of the present invention means the presence of the described features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or their groups. It should be understood that when we say that an element is "connected" or "coupled" to another element, it can be directly connected or coupled to other elements, or there may also be intermediate elements. In addition, the "connection" or "coupling" used herein may include wireless connection or coupling. The phrase "and / or" used herein includes any unit and all combinations of one or more related listed items.

[0056] Those skilled in the art can understand that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by those of ordinary skill in the art to which this invention belongs. It should also be understood that terms such as those defined in a general dictionary should be understood to have a meaning consistent with the meaning in the context of the prior art, and will not be interpreted in an idealized or overly formal sense unless defined as such here.

[0057] For ease of understanding the embodiments of the present invention, the following will further explain and illustrate with several specific embodiments in conjunction with the accompanying drawings, and each embodiment does not constitute a limitation on the embodiments of the present invention.

[0058] In view of the characteristics of the information flow in intelligent substations, the present invention combines the advantages of the TAS, SPQ, and Deficit Weight Round Robin (DWRR) algorithms, and designs a Dynamic Weighted Scheduling Algorithm with Time-Aware Shaping (DW-TAS). The proposed algorithm fully considers the characteristics of various communication services in intelligent substations and can better match the communication service requirements of intelligent substations.

[0059] The communication network of an intelligent substation undertakes various types of communication services, and the delay requirements and priorities of each service vary greatly, which poses many challenges for queue scheduling to meet different service requirements. The IEC 61850 series of standards are the general communication standards for digital substations, which define three main communication protocols, the MMS protocol, the GOOSE protocol, and the SV protocol. These protocols carry the main communication services of the substation and pose different requirements for the scheduling mechanism. In addition, the substation also uses time synchronization messages based on the Simple Network Time Protocol (SNTP). The above messages carry various types of service information in the substation and pose different requirements for queue scheduling. According to the different functions and delay requirements of the data flow, the traffic characteristics and delay requirements of the communication services in intelligent substations are shown in Table 1.

[0060] Table 1 Traffic Delay Requirements and Characteristics of Communication Services in Intelligent Substations

[0061]

[0062] (1) The trip command is an important fast message in the substation, which is used to cut off the circuit breaker in case of a fault and participate in functions of the substation such as relay protection and distance protection. The trip command affects the fault handling speed of the substation and has extremely high requirements for transmission delay. It is necessary to ensure that the communication delay does not exceed 3 ms. Moreover, the trip command will burst when a fault occurs and be transmitted intensively within a short period of time.

[0063] (2) The SV message, namely the sampled value digital transmission information, is generated after sampling the analog signals of current and voltage transformers. It belongs to the original data message and is mainly used to transmit the periodic sampled measurement values between the process layer electronic sensors and the bay-level devices. It is an important part of the communication between the process layer and the bay-level devices. The SV belongs to the periodic critical service and has the characteristics of large traffic, high real-time performance and high reliability.

[0064] (3) The switch position change information, like the trip command, is usually generated after a fault occurs and is used to record and notify the upper-level device of the current change in the switch state, mainly for substation event record analysis. This information has certain requirements for transmission delay, but the requirements are lower than those of the trip command. Usually, it is necessary to ensure that the transmission delay does not exceed 20 ms.

[0065] (4) The device status information is used to record the status information of various devices in the substation, such as transformers and circuit breakers. Through this information, the device status can be detected in real time and the fault can be located. The device status information message is a periodic data stream and has low requirements for transmission real-time performance. The transmission delay requirement does not exceed 100 ms.

[0066] (5) The time synchronization message is used to synchronize the clocks of each IED in the substation to ensure that the collaborative work between devices maintains a consistent time reference. The time synchronization message has relatively loose requirements for delay because its core goal is to maintain the symmetry of the message round-trip delay rather than strictly limit the transmission delay. In an intelligent substation, the symmetry of the message round-trip delay can be solved through mechanisms such as precise time stamps and path delay compensation. Therefore, the time synchronization message has relatively low requirements for delay. The IEC61850 standard does not currently clearly stipulate the delay requirements for the time synchronization message.

[0067] (6) The file transfer information mainly includes daily operation diaries, fault records and fault log files, etc. It has little impact on primary and secondary equipment and has no specific delay requirements. In case of conflicts, other service packets should be given priority.

[0068] By analyzing the delay requirements of substation communication services, the following special requirements for intelligent substation communication services in terms of dispatching can be summarized.

[0069] 1: Deterministic transmission of periodic sampled value messages.

[0070] Sampled value messages are typical periodic traffic in substation communication, used to transmit sampled values of current and voltage. They have extremely high requirements for latency and must ensure the deterministic transmission of messages.

[0071] 2: Loose scheduling and processing of file transfer information.

[0072] File transfer information is mainly used for record and device setting value transmission and has no specific latency requirements. Even if packets are lost for this type of information, they can be compensated through the retransmission mechanism. Therefore, the scheduling mechanism should allocate the lowest priority resources for such services to avoid interfering with high-priority services.

[0073] 3: Priority guarantee for bursty data.

[0074] Bursty data in intelligent substations includes tripping commands and switch position change information. These messages need to be transmitted immediately in case of a fault. Especially for tripping commands, their transmission delay is directly related to the stability of the power system. Therefore, the scheduling mechanism needs to minimize the transmission delay of bursty GOOSE messages while ensuring the transmission latency of SV messages.

[0075] The communication network traffic in intelligent substations has the characteristics of coexistence of periodicity and burstiness, which poses strict requirements on the scheduling algorithm. However, existing traditional scheduling algorithms are difficult to meet its complex requirements. Although SPQ can guarantee the priority scheduling of high-priority traffic, this algorithm will cause the "starvation" phenomenon of low-priority traffic; TAS realizes the deterministic transmission of periodic traffic through a time-triggered gated list, but it has insufficient support for aperiodic traffic and low resource utilization. Therefore, in view of the characteristics of the information flow in intelligent substations, the present invention combines the advantages of TAS, SPQ, and DWRR algorithms and designs a DW-TAS scheduling algorithm based on time-aware shaping. This algorithm provides fixed transmission time slots for periodic SV traffic through TAS technology to ensure the determinism of its transmission. At the same time, it combines the flexible scheduling of SPQ and DWRR algorithms for aperiodic traffic and preferentially processes bursty time-sensitive traffic through dynamic weighting to optimize its latency performance.

[0076] 2.1 Guarantee of determinism for periodic SV messages

[0077] In view of the characteristic of the deterministic transmission of periodic data streams such as SV messages in intelligent substations, the algorithm of the present invention uses the TAS scheduling mechanism to schedule SV messages separately. TAS uses the design of the gated list to control the opening and closing times of the queue, and the queue data can be forwarded only when the gated list corresponding to the queue is in the open state.

[0078] The algorithm of the present invention divides the communication time into periodic time periods by using a gating list, and further divides each period into two types of time slices, namely, periodic time slices and aperiodic time slices. After sorting and prioritizing the substation service traffic, where the SV message is a typical periodic traffic in the substation and is transmitted within the periodic time slice; the MMS, GOOSE, and time synchronization messages are transmitted within the aperiodic slice. The division of a communication time provided by an embodiment of the present invention is as Figure 1 shown.

[0079] The SV message records the digital sampling results of the voltage and current values in the substation. Under the 50Hz system used in our country, the sampling rate of the SV message is 4000Hz, and at this time, each sampling period is 250μs. Therefore, the gating list period is set to 250μs. After the nth cycle SV message enters the buffer queue, it is transmitted within the periodic time slice of cycle n. The remaining GOOSE messages, MMS messages, and time synchronization messages are transmitted within the aperiodic time slice. To ensure that the SV message can start transmission on time after arriving at the queue and is not affected by the data transmission within the aperiodic time slice, a guard band is set before the start of the periodic time slice, and the size of the guard band is determined according to the delay required for the Ethernet maximum transmission unit (MTU) data packet to be transmitted in the switch.

[0080] Let the length of the SV message be L SV bytes, and there are n messages passing through in each period. At this time, the size T of the periodic time slice is:

[0081]

[0082] In the formula: R is the link transmission rate. The present invention assumes that each application protocol data unit (APDU) in the message sent by the Merging Unit (MU) IED includes 4 application service data units (ASDUs), that is, it contains 4 voltage and current sampling data. There are 4 SV messages passing through the switch in each sampling period of the merging unit. At this time, the frame length of the SV message is 160Byte, that is:

[0083]

[0084] At this time, the size of the periodic time slice is:

[0085]

[0086] Loose scheduling and processing of file transfer information. For the file transfer information in an intelligent substation, since the file transfer information has little impact on primary equipment and secondary equipment and has no specific delay requirements, the present invention designs a nested structure of the DWRR algorithm and the SPQ algorithm according to this feature in the scheduling algorithm.

[0087] The principle of a TAS-PDWRR queue scheduling algorithm provided by an embodiment of the present invention is as Figure 2 shown. Q 1 ,Q 2 ,...,Q m represents m queues on a network node, corresponding to m types of traffic in the network respectively, and the network node allocates corresponding storage space for each queue. VQ 1 ,VQ 2 ,...,VQ m represents that a queue formed by rearranging several queues according to a certain scheduling rule is called a virtual queue. The algorithm model for non-periodic traffic in the DW-TAS algorithm is as shown in Equation (4).

[0088]

[0089] Among them, Q 1 caches file transfer information, which has non-real-time requirements. Q 2 ,Q 3 ,Q 4 ,Q 5 cache tripping commands, switch position change information, device status information, and time synchronization messages with real-time requirements respectively. VQ 1 is a virtual queue rearranged from Q 2 to Q 5 according to the DWRR scheduling rule; VQ is a virtual queue formed by rearranging VQ 1 and Q 1 and is the final output queue, where the priority of VQ 1 is higher than that of Q 1 . When there is no data packet waiting in Q 2 to Q 5 , the algorithm processes the file transfer information in the last queue according to the SPQ algorithm to ensure that it does not affect other high-priority data streams.

[0090] The bursty data streams in an intelligent substation include tripping information and switch position change information, and have extremely high requirements for delay, which is crucial for ensuring the safe and stable operation of the substation. In view of this feature, when the present invention's algorithm schedules the traffic of queues Q 2 to Q 5 , it improves on the basis of the traditional DWRR algorithm and introduces a weight dynamic adjustment strategy.

[0091] In the traditional DWRR algorithm, the output bandwidth is allocated among queues through weight values, and the weight values are fixed. Therefore, the proportion of bandwidth occupied by each queue is also fixed. However, when a primary device in a smart substation fails, protection action information and switch position change information will flood in the form of bursty traffic. At this time, the traditional DWRR algorithm may not be able to adjust the output bandwidth allocation of each queue due to the fixed weights. If a large fixed bandwidth is allocated for bursty traffic in advance during design, it will lead to waste of bandwidth resources in non-bursty situations. Therefore, according to the data traffic characteristics of smart substations, the present invention designs a dynamic weight adjustment strategy based on real-time load. The strategy proposed by the present invention dynamically adjusts the weight value of the queue by monitoring the total number of bytes of the packets to be sent in the queue in real time, so as to provide more bandwidth resources for bursty GOOSE data streams. The specific rules are as follows:

[0092] (1) When the total number of bytes of the packets in the queue is lower than the set lower limit MinByte i , the queue weight remains unchanged at the initial value;

[0093] (2) When the total number of bytes of the packets is higher than the set upper limit MaxByte i , the queue weight is adjusted to the preset maximum value to give priority to ensuring bursty traffic;

[0094] (3) When the total number of bytes of the packets is between the upper limit MaxByte i and the lower limit MinByte i , the weight value is dynamically adjusted according to the adjustment formula to make full use of bandwidth resources and ensure the timely transmission of bursty traffic. The adjustment formula is shown in Equation (5):

[0095]

[0096] where init_wt i is the initial weight value of queue i. max_wt i is the maximum weight of queue i. Byte i is the total number of bytes of the packets to be sent in the queue, MinByte i and MaxByte i are the set upper and lower limit values, and k is the weight adjustment coefficient of queue i.

[0097] To meet the delay requirements of different types of traffic in a smart substation, the present invention specifically designs the initial weights and dynamic adjustment strategies of each queue.

[0098] Q 4 ,Q 5The queues cache the tripping commands and the switch position change information respectively, which are extremely sensitive to latency, being 3ms and 20ms respectively, and their weights are dynamically adjusted according to Equation (5);

[0099] Q 3 The queue is used to cache the device status information. Although there is a latency requirement, it is lower than that of the first two types of data. Therefore, the weight adjustment rule of Q 3 is shown in Equation (6);

[0100] wt 3 = 1 - wt 4 -wt 5 -wt 2 (6)

[0101] Q 2 The queue stores the time synchronization messages, which have a high latency tolerance, with a small amount of data and a fixed weight value. Therefore, the queue weight of Q 2 is fixed at a low value and remains unchanged.

[0102] The initial weights of the queues are init_wt 5 = 0.2, init_wt 4 = 0.2, init_wt 3 = 0.5, init_wt 2 = 0.1, and for Q 4 ,Q 5 the weight adjustment coefficient k 4 = 1, k 5 = 3, corresponding to the maximum weights max_wt 5 = 0.5, max_wt 4 = 0.3 respectively. Therefore, the minimum weight of Q 3 is min_wt 3 = 0.1, and the weight of Q 2 is fixed at wt 2 = 0.1.

[0103] Within a polling cycle, the total number of bytes that all queues can schedule is Total byte , then the bandwidth share of each queue is its weight value, satisfying the following relationship:

[0104]

[0105] Queue Q i The number of bytes that can be sent within a polling cycle is SendByte i as shown in (8):

[0106] SendByte i = Totalbyte *wt i (8)

[0107] To meet the special requirements of intelligent substation communication services in terms of scheduling, the TAS-PDWRR algorithm proposed in the present invention combines the periodic scheduling of TAS, the priority mechanism of SPQ, and the improved DWRR dynamic weight adjustment strategy in design. The algorithm uses the TAS mechanism to allocate a fixed transmission bandwidth for SV messages to ensure the deterministic transmission of periodic data; by combining SPQ and DWRR, it provides loose scheduling for low-priority file transfer information and reduces its impact on other traffic; on the DWRR algorithm, through the dynamic weight adjustment strategy, it provides higher scheduling flexibility for bursty critical traffic.

[0108] In the previous part, the present invention gives corresponding scheduling strategies for different special requirements of intelligent substation communication services. On this basis, the present invention designs a dynamic weighted scheduling algorithm based on time-aware shaping.

[0109] The scheduling algorithm flow of a communication message transmission method for intelligent substations based on time-aware shaping provided by the present invention is as Figure 3 shown, including the following processing procedures:

[0110] After the data packet arrives, it enters the corresponding queue according to its type. Due to the function of TAS, the switch processes the periodic real-time data stream within the periodic time slice, corresponding to the SV message in the intelligent substation, and processes the data packets in the remaining queues within the aperiodic time slice. During the processing, when there are data packets waiting in queues 2 to 5, the scheduling algorithm uses the polling scheduling mechanism to schedule the data packets in these queues in turn, and calculates the waiting byte count of the data packets in each queue during each round of scheduling, and dynamically updates the weight according to formulas (5) and (6). After the update, the corresponding send byte count SendByte is calculated according to the updated weight i for data packet sending. When there are no data packets waiting in queues 2 to 5, the algorithm processes the file transfer information in the last queue according to the SPQ algorithm to ensure that it does not affect other high-priority data streams.

[0111] Numerical example simulation analysis. The simulation model structure of a D2-1 type intelligent substation provided by the embodiment of the present invention is as Figure 4 shown. In order to verify the effectiveness and feasibility of the DW-TAS scheduling algorithm in the intelligent substation communication network, the present invention refers to the D2-1 type intelligent substation architecture and its service flow for simulation analysis, as Figure 4As shown in the figure. The architecture of the intelligent substation can be divided into three layers: the station control layer, the bay layer, and the process layer. In the D2-1 type intelligent substation, there are a total of 9 bays, including 1 bus-coupler bay, 2 transformer bays, and 6 feeder bays. Each bay is equipped with multiple IED devices, including Breaker IED, P&C IED, and MU IED. These IED devices are connected to the core switch through the bay switch, and finally form a star topology structure.

[0112] The communication traffic flows in the intelligent substation include periodic data, bursty data, and low-priority data. Periodic data includes the SV messages transmitted from the MU IED to the P&C IED. The device status information is regularly sent from the Breaker IED and the P&C IED to the station control layer server, and the time synchronization messages are transmitted between the time synchronization system and all devices. Bursty data includes that when a fault occurs, the P&C IED will send a trip command to the Breaker IED and transmit the relevant information to the server. At this time, the switch position change information will also be transmitted to the server. In terms of non-real-time data transmission, the station control host sends file transfer information to the IED.

[0113] In order to evaluate the performance of the TAS-PDWRR algorithm in the simulation, the present invention abstracts the communication traffic flows of the intelligent substation into a Figure 5 star topology as shown in the figure. Among them, the core switch receives and schedules the data flows of the IED devices from each bay. FLOW1-6 are 6 data flows in the case of co-network transmission, and the initial settings of the corresponding service types, queues, traffic and other parameters are shown in Table 2.

[0114] The simulation environment of the present invention is MATLAB2018b, and different data models are adopted according to the characteristics of various types of information during the simulation process. The trip signal and the switch position change information belong to bursty data flows, and these data flows are modeled using the Poisson distribution to simulate their burstiness and uncertainty. The periodic data flows, including the sampled value messages, device status information, and time synchronization messages, are simulated using a periodic distribution. The file transfer information is represented as a random data flow, and packet groups are generated at a certain probability within any time period, and are simulated using the Poisson distribution. When determining the parameters, first calculate the average transmission interval of the data packets according to the traffic size and the data packet length. The periodic data flows are modeled according to the length of the transmitted data packets, the initial transmission time, and the average transmission interval. The data flows of the Poisson distribution are determined according to the transmission size of the messages and the parameter λ, where λ reflects the average arrival rate of the messages, and the average arrival interval of the messages is 1 / λ. Therefore, the value of λ is determined by the average transmission interval.

[0115] In the simulation, the data rate of the switch port is set to 100 Mbps, the link propagation delay is set to 0.5 μs, and the switch processing delay is set to 5 μs to approximate the delay characteristics of an actual network environment.

[0116] Table 2 Packet lengths, initial traffic, and distributions of various services in the simulation

[0117]

[0118]

[0119] In this simulation experiment, three comparison algorithm schemes are set up to verify the effectiveness of the scheduling strategy proposed in the present invention. The specific comparison schemes are as follows:

[0120] Comparison scheme 1: Divide the GOOSE information, sampled value information, device status information, and file transfer information into 4 priority levels, which are set to 7, 6, 5, and 1 respectively, and use the SPQ algorithm for queue scheduling.

[0121] Comparison scheme 2: The same priority settings as in comparison scheme 1. In the scheduling strategy, two-level scheduling is adopted. The trip command and switch position change information are preferentially scheduled using the SPQ algorithm, while other data packets are processed using the DWRR algorithm. This is abbreviated as the (HSGP, Hybrid Scheduling Algorithm with GOOSE Priority) algorithm in the text.

[0122] Comparison scheme 3: Based on the deficit weighted round-robin scheduling algorithm, the service division is supplemented, and the priorities of the switch position change information, sampled value message, and time synchronization information are set to 6, 5, and 3 respectively. In the scheduling strategy, a two-level scheduling different from that in comparison scheme 1 is adopted. The first five types of data are scheduled using the DWRR algorithm, while the file transfer information is scheduled using the SPQ algorithm, and the file transfer information can only be scheduled when other queues are empty. This is abbreviated as the HDWRR algorithm in the text.

[0123] Comparison scheme 4: Adopt the basic scheduling algorithm of the present invention, but do not introduce the dynamic weight adjustment mechanism, which is abbreviated as the TAS-HDWRR algorithm in the text.

[0124] In this simulation, the trip command and switch position change information are considered as bursty real-time data streams, and the ratio of the trip command traffic to the switch position change information traffic is set to 3:1. To simulate the burst situation of GOOSE message traffic under different loads, the GOOSE traffic is gradually increased until the total traffic passing through the core switch reaches 100% of the maximum data volume supported by the output port. Except for the traffic size of the switch position change information, other parameters are set according to Table 4. During this process, the performance of each scheduling algorithm is statistically analyzed under different load conditions.

[0125] In the simulation experiment of the present invention, the simulation duration is set to 1 s. Performance tests are carried out on the DW-TAS scheduling algorithm and three comparison algorithms, and the end-to-end delay of all data streams received by the station control host is obtained. Delay analysis is also carried out on the periodic traffic transmission effect and the bursty traffic scheduling effect.

[0126] When the bursty traffic of GOOSE messages increases and the load reaches 80%, the histogram of the delay distribution of SV message transmission for each algorithm is as Figure 6 shown. Figure 6 In the figure, the abscissa represents the end-to-end delay, with the unit of μs, and the ordinate represents the occurrence frequency of the corresponding delay. Figure 6 (a), Figure 6 (b), Figure 6 (c) and Figure 6 (d) show the SV message transmission performance of comparison schemes 1-3 and the scheme of the present invention under the condition of severe GOOSE traffic bursts. It can be seen from Figure 6 (a) that even under the SPQ algorithm, the delay distribution of SV messages is still significantly affected, showing a large delay jitter. Especially under the HDWRR algorithm, the delay distribution characteristics of SV messages deteriorate further, indicating that it cannot effectively guarantee the transmission of periodic messages when dealing with bursty traffic. However, Figure 6 (d) shows that when the algorithm of the present invention is adopted, the end-to-end transmission delay of SV messages always remains at 18.3 μs, without obvious jitter, realizing the deterministic transmission of periodic SV data streams.

[0127] Figure 7 is the box plot of the delay jitter of different scheduling algorithms under the condition of traffic bursts. It can be seen from Figure 7 that the algorithm of the present invention can still significantly reduce the delay jitter of SV messages even when the bursty traffic is large.

[0128] The bursty traffic scheduling effect Figures 8 - 10 respectively shows the average end-to-end delay of several algorithms for GOOSE message bursts at different levels. Figure 8 is the trip command, Figure 9 is the switch position change information, Figure 10 is the device status information.

[0129] From Figures 8 to 10 it can be observed that as the GOOSE bursty traffic increases, the end-to-end delay of various messages gradually increases with the load. For the transmission delay of the trip command, as Figure 8As shown, since the SPQ algorithm and the HSGP algorithm preferentially schedule tripping command messages, they can maintain the lowest average delay under burst traffic conditions. The transmission delay of tripping commands under the HDWRR algorithm is better than that under the algorithm of the present invention and the TAS-HDWRR algorithm. This is because after the TAS scheduling algorithm is introduced in the algorithm of the present invention and the TAS-HDWRR algorithm, fixed transmission time slots are reserved for SV messages. When other types of messages arrive during the SV message transmission time slot, they need to wait for the next available message transmission time slot, resulting in a certain increase in delay. In addition, in Figure 8 when the switch traffic reaches 40 Mbps, the algorithm of the present invention optimizes the transmission efficiency of tripping commands by dynamically adjusting the strategy, and its delay performance is better than that of the TAS-HDWRR algorithm.

[0130] Figure 9 shows the end-to-end delay distribution of switch position change information. Consistent with the conclusion of Figure 8 , the transmission delay of switch position change information under the HDWRR algorithm is still lower than that under the algorithm of the present invention and the TAS-DWRR algorithm. As the traffic burst causes the total switch traffic to exceed 40 Mbps, compared with the TAS-DWRR algorithm, the algorithm of the present invention effectively reduces the transmission delay of switch position change information by virtue of the dynamic weight adjustment mechanism.

[0131] Figure 10 is the end-to-end delay distribution of device status information. It can be observed that the delay of device status information under the HDWRR algorithm is better than that under the algorithm of the present invention and the TAS-HDWRR algorithm, which is also caused by TAS reserving bandwidth for SV messages. In addition, when the traffic burst causes the total switch traffic to exceed 40 Mbps, the transmission delay of device status information under the algorithm of the present invention is higher than that of the TAS-HDWRR algorithm. This is because when the algorithm of the present invention faces a large number of burst GOOSE messages, more bandwidth is allocated to tripping commands and switch position change information through dynamic weight adjustment, resulting in a reduction in the bandwidth of device status information. However, even after the weight adjustment, the device status information can still obtain a fixed minimum bandwidth under the polling scheduling mechanism, thus avoiding performance degradation caused by traffic accumulation. Considering that the delay requirement of device status information is relatively looser than that of tripping commands and switch position change information, under extreme load conditions, the transmission delay of device status information caused by this algorithm strategy is still within an acceptable range.

[0132] In summary, in the embodiments of the present invention, all communication services are transmitted over the same network in an intelligent substation, increasing the risk of high latency caused by network congestion. To meet the latency requirements of multiple services, in view of the traffic characteristics of coexisting periodic traffic and bursty traffic in the communication network of the intelligent substation, a dynamic weighted scheduling algorithm based on time-aware shaping is designed. Simulation experiments show that the algorithm of the present invention can, under bursty load conditions, transmit more critical bursty GOOSE traffic as much as possible on the premise of ensuring the basic transmission requirements of less critical traffic, and at the same time can achieve its deterministic transmission by allocating fixed transmission time slots for periodic SV traffic.

[0133] Those of ordinary skill in the art can understand that: The drawings are only schematic diagrams of one embodiment, and the modules or processes in the drawings are not necessarily essential for implementing the present invention.

[0134] From the description of the above embodiments, those skilled in the art can clearly understand that the present invention can be implemented by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present invention, in essence, or the part that makes a contribution to the prior art, can be embodied in the form of a software product. The computer software product can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in various embodiments or some parts of the embodiments of the present invention.

[0135] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments. In particular, for the device or system embodiments, since they are basically similar to the method embodiments, they are described relatively simply, and the relevant parts can refer to the partial description of the method embodiments. The device and system embodiments described above are only illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they can be located in one place, or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative work.

[0136] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed by the present invention should be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A method for transmitting communication messages in a smart substation based on time-aware shaping, characterized in that: include: Allocate a fixed periodic time slice to the periodic sampling value SV message in the smart substation through time-aware shaping TAS, and transmit the SV message within the periodic time slice; The weight value of the queue is dynamically adjusted by real-time monitoring of the total number of bytes of non-periodic service messages to be sent in the queue. According to the weight value of each queue, the dynamic weighted DW-TAS scheduling algorithm based on time-aware shaping is used to schedule the transmission of non-periodic service messages in the queue.

2. The method according to claim 1, characterized in that The method of allocating a fixed periodic time slice to a periodic sampling value SV message in a smart substation through time-aware shaping TAS, and transmitting the SV message within the periodic time slice, includes: The communication time is divided into periodic time slices and non-periodic time slices using a gating list. After the nth periodic SV message enters the buffer queue, it is transmitted in the periodic time slice of period n, and the remaining non-periodic service messages are transmitted in the non-periodic time slice. A protection band is set before the start of the periodic time slice. The size of the protection band is determined based on the delay required for the Ethernet maximum transmission unit data packet to be transmitted in the switch.

3. The method according to claim 1, characterized in that The method of allocating a fixed periodic time slice for a periodic SV message through the TAS and transmitting the SV message within the periodic time slice includes: Assume the length of SV message is L SV Bytes, n messages pass through each cycle, then the size T of the periodic time slice is: Where: R is the link transmission rate; Assume that each application protocol data APDU in the message includes 4 application service data units ASDU, that is, it contains 4 voltage and current sampling data, and 4 SV messages pass through the switch in each sampling cycle of the merging unit. At this time, the frame length of the SV message is 160Byte, that is: At this time, the size of the periodic time slice is: 。 4. The method according to claim 1, characterized in that: The method of dynamically adjusting the weight value of the queue by real-time monitoring the total number of bytes of non-periodic service messages to be sent in the queue includes: Set Q1, Q2, ..., Q m It represents m queues on the network node, corresponding to m types of traffic in the network. Q1 caches file transfer information with non-real-time requirements. Q2, Q3, Q4, and Q5 cache trip commands, switch position change information, device status information, and time synchronization messages with real-time requirements, respectively. The total number of bytes of messages to be sent in the queue is monitored in real time to dynamically adjust the queue weight. The specific rules are as follows: (1) When the total number of bytes in the queue is lower than the set lower limit MinByte i When , the queue weight remains unchanged at the initial value; (2) When the total number of bytes in the message exceeds the set upper limit MaxByte i When the queue weight is adjusted to the preset maximum value, the burst traffic is given priority. (3) When the total number of bytes in the message is between the upper limit MaxByte i and the lower limit MinByte i When the weight value is between , the weight value is dynamically adjusted according to the adjustment formula to make full use of bandwidth resources and ensure the timely transmission of burst traffic. The adjustment formula is shown in formula (5): Among them, init_wt i is the initial weight value of queue i, max_wt i is the maximum weight of queue i, Byte i MinByte is the total number of bytes of messages to be sent in the queue. i and MaxByte i are the upper and lower limits set, k is the weight adjustment coefficient of queue i; The weights of the Q4 and Q5 queues are dynamically adjusted according to formula (5), and the weight adjustment rule of Q3 is shown in formula (6); wt3=1-wt4-wt5-wt2 (6) The queue initialization weights are init_wt5=0.2, init_wt4=0.2, init_wt3=0.5, init_wt2=0.1, and the weight adjustment coefficients of Q4 and Q5 are k4=1 and k5=3, corresponding to the maximum weights max_wt5=0.5 and max_wt4=0.3 respectively. The minimum weight of Q3 is min_wt3=0.1, and the weight of Q2 is fixed at wt2=0.

1.

5. The method according to claim 4, characterized in that The method of scheduling and transmitting the non-periodic service messages in the queues by using the dynamic weighted DW-TAS scheduling algorithm based on time-aware shaping according to the weight values ​​of each queue includes: The scheduling algorithm for non-periodic message traffic in the dynamic weighted DW-TAS scheduling algorithm is shown in formula (4). VQ1 is a virtual queue rearranged by Q2 to Q5 according to the DWRR scheduling rule. VQ is a virtual queue formed by rearranging VQ1 and Q1 according to the SPQ scheduling rule. It is the final output queue. The priority of VQ1 is higher than that of Q1. The total number of bytes that can be scheduled by all queues in one polling cycle is Total byte , then the bandwidth proportion of each queue is its weight value, satisfying the following relationship: Queue Q i The number of bytes that can be sent in one polling cycle SendByte i As (8) indicates: SendByte i =Total byte *wt i (8) After the message arrives, it enters the corresponding queue according to its type. In the periodic time slice, the switch transmits the SV message. In the non-periodic time slice, when there are data packets waiting in Q2~Q5, the scheduling algorithm uses the polling scheduling mechanism to schedule the data packets in Q2~Q5 in turn, and calculates the number of waiting bytes of the data packets in each queue in each round of scheduling, and dynamically updates the weight according to formula (5) and formula (6), and calculates the corresponding number of sent bytes SendByte according to the updated weight. i , according to SendByte i When there are no data packets waiting in queues 2 to 5, the algorithm processes the file transfer information in the last queue according to the SPQ algorithm.

Images