Multi-switch TSN dynamic scheduling method and system based on fuzzy control, and electronic equipment

Abstract

The invention provides a multi-switch TSN dynamic scheduling method and system based on fuzzy control and electronic equipment, and the method comprises the steps: obtaining the end-to-end time delay of a time-sensitive flow when the time-sensitive flow is forwarded by a switch through a topology discovery module, outputting an adjustment coefficient k based on a TSN fuzzy controller, configuring a gating list time-sensitive window value of a TSN switch according to a fuzzy rule, and obtaining the time-sensitive window value of the time-sensitive flow when the time-sensitive flow is forwarded by the switch; and the gating list is configured into the TSN interaction machine through the network configuration module, so that the average time delay satisfaction rate of the time-sensitive flow is improved, and better flexibility is achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of industrial communication network, more particularly to a multi-switch time sensitive network (TSN) dynamic flow scheduling method and system based on fuzzy control and electronic equipment. BACKGROUND

[0002] With the rapid development of industrial internet, the demand for deterministic network increases. Time sensitive network (TSN) is deployed in large scenarios such as smart factory and smart grid. In order to meet the needs of large-scale access of multiple devices and a large number of nodes, a large number of TSN switches are deployed, resulting in a large number of TSN switches in the middle. In the network topology, the periodic time sensitive flow loses periodicity after being shaped by the switch directly connected to it. The traditional edge scheduling algorithm and adaptive time sensitive window (ATSW) cannot be applied to the scene of a large number of TSN switches, which means that the ATSW algorithm is only applicable to the TSN switch directly connected to the time sensitive flow sending terminal.

[0003] In order to solve the problem that the traditional ATSW algorithm is only applicable to the switch directly connected to the time sensitive flow source node, and draw lessons from the control theory in industrial production, since the time delay of time sensitive flow and the size of time sensitive window are not in linear relationship, the commonly used PID control is not applicable, and a more flexible dynamic scheduling algorithm is needed. SUMMARY

[0004] The present application provides a multi-switch time sensitive network (TSN) dynamic flow scheduling method and system based on fuzzy control and electronic equipment, a fuzzy control-based adaptive time sensitive window dynamic scheduling algorithm FC-ATSW, which configures the GCL (gate list) time sensitive window value of the intermediate switch according to a series of fuzzy rules, so that the average time delay of time sensitive flow is improved.

[0005] The first aspect of the present application discloses a multi-switch TSN dynamic scheduling method based on fuzzy control, applied in a multi-switch network environment, the method comprising:

[0006] Step S1: acquiring the end-to-end time delay L(t) of time sensitive flow when forwarded by TSN switch through a topology discovery module;

[0007] Step S2: obtaining a worst delay upper limit L acceptable to a user when the time-sensitive flow is successfully forwarded in the TSN switch exp , the L exp As the input of the TSN fuzzy controller, the error D(t) of the actual delay of the time-sensitive flow and the acceptable worst delay upper limit, the change ΔD(t) of the error, and the size ΔT2(t) of the last time-sensitive window adjustment are used to control the adjustment coefficient k, the TSN fuzzy controller outputs the adjustment coefficient k, and the TSN fuzzy controller makes the real-time delay of the time-sensitive flow less than the acceptable worst delay upper limit L of the user exp , and infinitely close to L exp ;

[0008] Step S3: adjusting the time-sensitive window size based on the obtained adjustment coefficient k to obtain a gating list of the TSN switch;

[0009] Step S4: configuring the gating list into the TSN switch through a network configuration module.

[0010] According to the multi-switch TSN dynamic scheduling method based on fuzzy control of the first aspect of the present application, the fuzzy controller in step S2 includes D(t), ΔD(t), and ΔT2(t), specifically:

[0011] The error D(t) of the actual delay of the time-sensitive flow and the acceptable worst delay upper limit is:

[0012]

[0013] The change ΔD(t) of the error is the change value of the normalized error D(t):

[0014] ΔD(t) = D(t) - D(t-1);

[0015] The size ΔT2(t) of the last time-sensitive window adjustment is the normalized change value of the time-sensitive window size T2:

[0016]

[0017] T is the period length of the gating list of the TSN switch.

[0018] According to the multi-switch TSN dynamic scheduling method based on fuzzy control of the first aspect of the present application, further comprising:

[0019] The D(t) can be used to control the direction of the time-sensitive window adjustment, and D(t), ΔD(t), and ΔT2(t) jointly control the speed of the adjustment.

[0020] According to the multi-switch TSN dynamic scheduling method based on fuzzy control, the step S3 further comprises:

[0021] After the adjustment coefficient k is obtained, the size of the time sensitive window at the next time is calculated:

[0022]

[0023] After the updated T2 is obtained, a threshold list period length calculation formula is obtained:

[0024] T=T1+T2+T3;

[0025] wherein T3 is a low priority window, T1 is a guard band length, and l eth_max and an egress bandwidth B out determines:

[0026]

[0027] According to the multi-switch TSN dynamic scheduling method based on fuzzy control, the method further comprises:

[0028] The membership function of D(t) is established, the domain of the input quantity D(t) is (-∞, +∞), and five membership functions are defined in turn as follows:

[0029]

[0030]

[0031] The membership function of ΔD(t) is established, the domain of the input quantity ΔD(t) is also (-∞, +∞), the membership functions are selected in a left-right symmetric form, and five membership functions are defined in turn as follows:

[0032]

[0033]

[0034] The membership function of ΔT2(t) is established, the domain of the input quantity ΔT2(t) is (-1, +1), and five membership functions are defined in turn as follows:

[0035]

[0036] According to the multi-switch TSN dynamic scheduling method based on fuzzy control, ΔT2(t) reflects the change of the time sensitive window T2, that is, the change of T2 when the dynamic configuration of the TSN switch is performed last time.

[0037] According to the method for dynamically scheduling multiple switches based on fuzzy control, the fuzzy control rule of the fuzzy controller is specifically:

[0038] When the current time-sensitive flow delay is less than the acceptable worst delay upper limit, the time-sensitive window value is reduced;

[0039] When the current time-sensitive flow delay is equal to the acceptable worst delay upper limit, the time-sensitive window does not need to be adjusted any more;

[0040] When the current time-sensitive flow delay exceeds the acceptable worst delay upper limit, the time-sensitive window is increased.

[0041] The second aspect of the present application discloses a system for dynamically scheduling multiple switches based on fuzzy control, which comprises a processing unit configured to perform the steps of the method for dynamically scheduling multiple switches based on fuzzy control of the first aspect.

[0042] The third aspect of the present application discloses an electronic device comprising a memory and a processor, wherein the memory stores a computer program, and the processor is configured to execute the program in the memory to implement the method for dynamically scheduling multiple switches based on fuzzy control of the first aspect.

[0043] The fourth aspect of the present application discloses a computer-readable storage medium storing the method for dynamically scheduling multiple switches based on fuzzy control of the first aspect.

[0044] In summary, the present application provides a method and system for dynamically scheduling multiple switches based on fuzzy control, which obtains the end-to-end delay of time-sensitive flow when forwarded by the switch through a topology discovery module, adjusts the coefficient k based on the output of the TSN fuzzy controller, configures the time-sensitive window value of the gate list of the TSN switch according to the fuzzy rule, and configures the gate list into the TSN switch through a network configuration module, so that the average delay satisfaction rate of the time-sensitive flow is improved, and the flexibility is better. BRIEF DESCRIPTION OF DRAWINGS

[0045] In order to more clearly illustrate the specific embodiments of the present application or the technical solutions in the prior art, the drawings needed in the specific embodiments or prior art description will be briefly introduced below. Obviously, the drawings in the following description are some embodiments of the present application, and those skilled in the art can obtain other drawings according to these drawings without creative labor.

[0046] Figure 1 is the flow chart of the method for dynamically scheduling multiple switches based on fuzzy control of the present application;

[0047] Figure 2 is a structure diagram of a TSN fuzzy controller of the present application;

[0048] Figure 3 is a membership function curve of D(t) of the present application;

[0049] Figure 4 is a membership function curve of AD(t) of the present application;

[0050] Figure 5 is a membership function curve of AT2(t) of the present application;

[0051] Figure 6 is a membership function curve of the adjustment coefficient k of the present application. DETAILED DESCRIPTION

[0052] In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0053] In the TSN fuzzy control system, the channel transmission model is a three-section transmission mode. It includes a guard band, a time-sensitive window and a low-priority window. The above three-section transmission mode is used as a period for the gate scheduling to be alternately executed. The GCL value of the guard band is "CCCCCCCC", at which time all the traffic will be waiting in the queue. Its function is to ensure the transmission of low-priority traffic, without affecting the transmission of time-sensitive traffic. The GCL value of the time-sensitive window is "oCCCCCCC", through which the traffic with a priority of 7 can pass, and other traffic enters the queue to wait. The time-sensitive window can ensure that only time-sensitive flow can pass within the GCL time, without other low-priority traffic competing with it. The GCL value of the low-priority window is "Cooooooo", through which the traffic with a priority of 0-6 can pass, and the low-priority traffic is scheduled according to strict priority.

[0054] The present application proposes a dynamic traffic scheduling method in a time-sensitive network (TSN). The framework of the scheduling method is as follows: a topology discovery module obtains the end-to-end delay of time-sensitive flow when the switch forwards, L expAs the upper limit of the worst delay acceptable to the user when the switch can successfully forward the time-sensitive flow, the value is taken as the input of the TSN fuzzy controller together with a series of values obtained by the topology discovery module, the TSN fuzzy controller adopts a three-dimensional structure, and three input variables are: D(t) is the normalized error of the actual delay of the time-sensitive Ethernet frame and the acceptable upper limit of the worst delay L exp L(t) is the worst actual delay of the time-sensitive Ethernet frame in the current period. For the convenience of designing the membership function, D(t) is the normalized result.

[0055] The control target of the TSN fuzzy controller is to make the real-time delay of the time-sensitive flow less than the upper limit of the worst delay L exp acceptable to the user and as close to L exp as possible to avoid the time-sensitive flow occupying too much bandwidth and affecting the transmission of low-priority flows. Among the three input variables, D(t) can be used to control the direction of time-sensitive window adjustment, D(t), AD(t) and AT2(t) jointly control the speed of adjustment. The adjustment coefficient k is the output quantity of the fuzzy controller, and the size of the adjustment coefficient k is controlled by establishing membership functions for the three input quantities through fuzzy control to further adjust the GCL value. Finally, the updated GCL value is obtained and configured on the TSN switch.

[0056] As shown in Figure 1 , the application provides a multi-switch TSN dynamic scheduling method based on fuzzy control, comprising the following steps:

[0057] Step S1: obtaining the end-to-end delay L(t) of the time-sensitive flow when forwarded by the TSN switch through the topology discovery module; according to the TSN fuzzy control system, the topology discovery module first obtains the end-to-end delay L(t) of the time-sensitive flow when forwarded by the switch.

[0058] Step S2: obtaining the upper limit L of the worst delay acceptable to the user when the time-sensitive flow is successfully forwarded in the TSN switch exp , taking L(t), L exp as the input of the TSN fuzzy controller, controlling the adjustment coefficient k according to the error D(t) of the actual delay of the time-sensitive flow and the acceptable upper limit of the worst delay, the change AD(t) of the error and the size AT2(t) of the last time-sensitive window adjustment, the TSN fuzzy controller outputs the adjustment coefficient k, and the TSN fuzzy controller makes the real-time delay of the time-sensitive flow less than the upper limit L exp of the worst delay acceptable to the user and infinitely close to L exp .

[0059] As shown in Figure 2 , the actual delay of the time-sensitive Ethernet frame and the acceptable upper limit of the worst delay L expThe normalized error D(t) is:

[0060] The ΔD(t) is defined as the change of the normalized error D(t):

[0061] ΔD(t) = D(t) - D(t-1)

[0062] The ΔT2(t) is defined as the normalized change of the time sensitive window size T2:

[0063]

[0064] T is the period length of the GCL of the controlled TSN switch.

[0065] Next, the three input variables can be used to construct a three-dimensional fuzzy controller. Among the three input variables, D(t) can be used to control the direction of time sensitive window adjustment, and D(t), ΔD(t) and ΔT2(t) together control the speed of adjustment. The adjustment coefficient k is the output of the fuzzy controller.

[0066] As shown in FIG. 1, the membership function curves of the three input variables are constructed. Figures 3-4

[0067] The domain of the input D(t) is (-∞, +∞), and the fuzzy set is defined as {NB, NS, ZO, PS, PB}, NB is negative big, NS is negative small, ZO is zero, PS is positive small, and PB is positive big. The GCL values are configured for multiple switches.

[0068] The five membership functions used are defined as follows:

[0069]

[0070]

[0071] The domain of the input ΔD(t) is also (-∞, +∞), and the fuzzy set is defined as {NB, NS, ZO, PS, PB}. The left and right symmetric forms are used to select the membership function. The five membership functions used are defined as follows:

[0072]

[0073]

[0074] ​Establishment of membership function of ΔT2(t). The domain of input quantity ΔT2(t) is (-1, +1), and its fuzzy set is defined as {NB, NS, ZO, PS, PB}, and the five membership functions are defined as follows:

[0075]

[0076] ΔT2(t) reflects the change of time sensitive window T2, i.e. the change of T2 when the GCL of TSN switch is dynamically configured last time, as shown in Figure 5 .

[0077] The curve of membership function of adjustment coefficient k is shown in Figure 6 , and the domain of output quantity k is (-1, +1). The fuzzy set is divided more finely than the input quantity, and is defined as {NB, NM, NS, ZO, PS, PM, PB}. The membership function is selected as single value type.

[0078] The size of adjustment coefficient k is controlled according to the error D(t) of actual time delay of time sensitive flow and acceptable upper limit of worst time delay, the change ΔD(t) of error and the size ΔT2(t) of last time sensitive window adjustment, and the GCL value of TSN switch is further adjusted. The following cases are considered to establish fuzzy control rules and use if-then statement to realize, and each case corresponds to a fuzzy statement.

[0079] Rule1: IF D(t) is NB, AND ΔD(t) is NB, THEN k is NB; when the current time sensitive flow delay is much smaller than the acceptable upper limit of worst time delay, and the error change is negative large, i.e. the current time sensitive window is too large, and the current time sensitive flow delay is much smaller than the delay at last time, the possible situation is that the time sensitive traffic is less, so at this time, in order to avoid bandwidth loss, the time sensitive flow window should be greatly reduced.

[0080] Rule2: IF D(t) is NB, AND ΔD(t) is NS, THEN k is NM; when the current time sensitive flow delay is much smaller than the acceptable upper limit of worst time delay, and the error change is negative small, at this time, the situation is similar to Rule1, and the difference is that the current time sensitive flow delay is slightly smaller than the delay at last time, at this time, the time sensitive flow window should be moderately reduced.

[0081] Rule 3: IF D(t) is NB, AND AD(t) is ZO or PS, THEN k is NB; when the current time sensitive flow delay is far less than the acceptable worst delay upper bound, and the error change is zero or positive, the situation may be that: the time sensitive window has been reduced at the last time, but the adjustment is small, and the effect is not obvious, so the time sensitive window value should be greatly reduced.

[0082] Rule 4: IF D(t) is NB, AND AD(t) is PB, THEN k is NM; when the current time sensitive flow delay is far less than the acceptable worst delay upper bound, and the error change is positive, the situation may be that: the time sensitive window has been reduced at the last time, and the effect is obvious, but in order to prevent the time sensitive window from being reduced too much at the next time, the time sensitive window value should be moderately reduced.

[0083] Rule 5: IF D(t) is NS, THEN k is NS; when the current time sensitive flow delay is slightly less than the acceptable worst delay upper bound, the adjustment of the time sensitive window should be more conservative, so the time sensitive window value should be slightly reduced.

[0084] Rule 6: IF D(t) is ZO, THEN k is ZO; when the current time sensitive flow delay is approximately equal to the acceptable worst delay upper bound, the GCL value of the TSN switch is already the correct value, and there is no need to adjust it.

[0085] Rule 7: IF D(t) is PS, AND AD(t) is NB, AND AT2(t) is PS, THEN k is PS; when the current time sensitive flow delay is slightly greater than the acceptable worst delay upper bound, and the error change is negative, and the time sensitive window increased slightly at the last time, the possible situation is that: the time sensitive flow delay at the last time exceeded the acceptable worst delay upper bound, and the error change effect was obvious after the time sensitive window was slightly increased, so the adjustment scheme of the last time is followed, that is, the time sensitive window is slightly increased.

[0086] Rule 8: IF D(t) is PS, AND AD(t) is NB, AND AT2(t) is PB, THEN k is PM; when the current time sensitive flow delay is slightly greater than the acceptable worst delay upper bound, and the error change is negative, and the time sensitive window increased a lot at the last time, the situation is similar to Rule 7, and the adjustment scheme of the last time is followed, that is, the time sensitive window is moderately increased.

[0087] Rule 9: IF D(t) is PS, AND AD(t) is NS or ZO, THEN k is PM; when the current time-sensitive flow delay is slightly beyond the acceptable upper bound of the worst delay, and the error change is negative small or zero, it indicates that the effect of the last time-sensitive window adjustment is small, so the time-sensitive window is increased in this adjustment.

[0088] Rule 10: IF D(t) is PS, AND AD(t) is PS, AND AT2(t) is NB, THEN k is PM; when the current time-sensitive flow delay is slightly beyond the acceptable upper bound of the worst delay, and the error change is positive small, and the time-sensitive window is greatly reduced at the last time, the possible situation is that the time-sensitive window is greatly reduced at the last time, which leads to the current time-sensitive flow delay not meeting the requirements, so the time-sensitive window should be moderately increased.

[0089] Rule 11: IF D(t) is PS, AND AD(t) is PS, AND AT2(t) is NS, THEN k is PS; when the current time-sensitive flow delay is slightly beyond the acceptable upper bound of the worst delay, and the error change is positive small, and the time-sensitive window is slightly reduced at the last time, the situation at this time is similar to Rule 10, and only the time-sensitive window value needs to be slightly increased.

[0090] Rule 12: IF D(t) is PS, AND AD(t) is PS, AND AT2(t) is ZO, THEN k is PS; when the current time-sensitive flow delay is slightly beyond the acceptable upper bound of the worst delay, and the error change is positive small, and the time-sensitive window is unchanged at the last time, the possible situation is that the GCL value of the TSN switch at the last time is in a suitable state, but the time-sensitive traffic increases at the current time, and the TSN switch is not enough to maintain the network demand, so the time-sensitive window needs to be slightly increased.

[0091] Rule 13: IF D(t) is PS, AND AD(t) is PS, AND AT2(t) is P, THEN k is PM; wherein, P represents PS and PB. When the current time-sensitive flow delay is slightly beyond the acceptable upper bound of the worst delay, and the error change is positive small, and the time-sensitive window is increased at the last time, the possible situation is that the time-sensitive flow delay has exceeded the acceptable worst delay at the last time, and the time-sensitive window has been expanded and adjusted, but the expansion degree is not enough, which leads to the further increase of the time-sensitive flow delay. Therefore, the time-sensitive window value should be moderately increased.

[0092] Rule 14: IF D(t) is PS, AND AD(t) is PB, AND AT2(t) is N, THEN k is PM; wherein N represents including NS and NB. When the current time sensitive flow delay slightly exceeds the acceptable worst delay upper bound, and the error change is positive, and the time sensitive window is reduced at the last time, similar to the case of Rule 10, the time sensitive window should be increased.

[0093] Rule 15: IF D(t) is PS, AND AD(t) is PB, AND AT2(t) is ZO, THEN k is PB; when the current time sensitive flow delay slightly exceeds the acceptable worst delay upper bound, and the error change is positive, and the time sensitive window is unchanged at the last time, the case is similar to Rule 12, and the time sensitive window should be increased.

[0094] Rule 16: IF D(t) is PS, AND AD(t) is PB, AND AT2(t) is P, THEN k is PB; when the current time sensitive flow delay slightly exceeds the acceptable worst delay upper bound, and the error change is positive, and the time sensitive window is greatly increased at the last time, the case is similar to Rule 13, and the time sensitive window should be increased.

[0095] Rule 17: IF D(t) is PB, THEN k is PB; when the current time sensitive flow delay far exceeds the acceptable worst delay upper bound, the case is relatively urgent, and there is no need to consider other factors, and the time sensitive window can be directly increased.

[0096] After obtaining the adjustment coefficient k, the size of the time sensitive window at the next time is calculated:

[0097]

[0098] After obtaining the new T2, the GCL cycle length calculation formula is obtained:

[0099] T = T1 + T2 + T3

[0100] Wherein T3 is a low priority window, T1 is a guard band length, which is determined by the maximum Ethernet frame size leth_max and the egress bandwidth, to ensure that a maximum Ethernet frame can be transmitted, and the guard band length is generally a fixed value:

[0101]

[0102] Step S3: adjusting the time sensitive window size based on the obtained adjustment coefficient k, to obtain the gate list of the TSN switch;

[0103] Step S4: configuring the gating list into the TSN switch by the network configuration module.

[0104] The application further provides a multi-switch TSN dynamic scheduling system based on fuzzy control.

[0105] The application further provides an electronic device comprising a memory and a processor, wherein the memory stores a computer program, and the processor is configured to execute the program in the memory to implement the multi-switch TSN dynamic scheduling method based on fuzzy control.

[0106] The application further provides a computer readable storage medium, which stores a computer program for implementing the multi-switch TSN dynamic scheduling method based on fuzzy control.

[0107] In summary, the multi-switch TSN dynamic scheduling method based on fuzzy control has the adaptive time-sensitive wide-port dynamic scheduling algorithm based on fuzzy control, analyzes the end-to-end delay during forwarding, inputs the TSN fuzzy controller, configures the GCL time-sensitive window value of the TSN switch based on a series of fuzzy rules by the fuzzy controller, configures the adjusted GCL to the TSN switch, and thus the average delay satisfaction rate of the time-sensitive flow is improved, and better flexibility is achieved.

[0108] Please note that the technical features of the above embodiments can be combined in any manner, and to make the description concise, not all possible combinations of the technical features in the above embodiments are described, however, as long as the combinations of the technical features do not exist, they should be considered as the scope of the description. The above embodiments only express several embodiments of the application, and the description is more specific and detailed, but it should not be understood as a limitation on the scope of the patent. It should be noted that for ordinary skilled in the art, without departing from the concept of the application, some modifications and improvements can be made, which are all within the scope of the application. Therefore, the patent protection scope of the application should be subject to the appended claims.

Claims

1. A method for multi-switch TSN dynamic scheduling based on fuzzy control, applied in a multi-switch network environment, characterized in that, The method comprises: Step S1: obtaining the end-to-end delay L(t) of the time-sensitive flow when forwarded by the TSN switch through a topology discovery module; Step S2: obtaining a worst delay upper limit L acceptable to users when the time-sensitive flow is successfully forwarded in the TSN switch exp , the L exp As the input of the TSN fuzzy controller, the adjustment coefficient k is controlled according to the error D(t) of the actual delay of the time-sensitive flow and the acceptable worst delay upper limit, the change ΔD(t) of the error and the size ΔT2(t) of the last time-sensitive window adjustment, the TSN fuzzy controller outputs the adjustment coefficient k, and the TSN fuzzy controller makes the real-time delay of the time-sensitive flow less than the acceptable worst delay upper limit L of the users exp , and infinitely close to L exp ; Step S3: adjusting the time-sensitive window size based on the obtained adjustment coefficient k to obtain the gating list of the TSN switch; Step S4: configuring the gating list into the TSN switch through a network configuration module.

2. The method of claim 1, wherein the method further comprises: The fuzzy controller in step S2 comprises D(t), ΔD(t) and ΔT2(t), specifically: The error D(t) of the actual delay of the time-sensitive flow and the acceptable upper limit of the worst delay is: The change ΔD(t) of the error is the change value of the normalized error D(t): ΔD(t) = D(t) - D(t-1); The size ΔT2(t) of the last time-sensitive window adjustment is the normalized change value of the time-sensitive window size T2: T is the period length of the gating list of the TSN switch.

3. The method of claim 2, wherein the method further comprises: Further comprising: D(t) can be used to control the direction of time-sensitive window adjustment, and D(t), ΔD(t) and ΔT2(t) jointly control the speed of adjustment.

4. The method of claim 1, wherein the method further comprises: The step S3 further comprises: After obtaining the adjustment coefficient k, the size of the time-sensitive window at the next moment is calculated: After obtaining the updated T2, the gating list period length calculation formula is obtained: T = T1 + T2 + T3; where T3 is a low priority window, T1 is a guard band duration, and l is the maximum Ethernet frame size eth_max and the egress bandwidth B out Decide:

5. The method of claim 3, wherein the method further comprises: Further comprising: The establishment of the D(t) membership function, the domain of the input D(t) is (-∞, +∞), and the five membership functions are defined in turn as follows: The establishment of the ΔD(t) membership function, the domain of the input ΔD(t) is also (-∞, +∞), the selection of the membership function adopts the form of left-right symmetry, and the five membership functions are defined in turn as follows: The establishment of the ΔT2(t) membership function, the domain of the input ΔT2(t) is (-1, +1), and the five membership functions are defined in turn as follows:

6. The fuzzy control-based multi-switch TSN dynamic scheduling method according to claim 5, characterized in that: ΔT2(t) reflects the change of the time-sensitive window T2, that is, the change of T2 when the gating list of the TSN switch is dynamically configured last time.

7. The method of claim 1, wherein the method further comprises: The fuzzy control rules of the TSN fuzzy controller are specifically: When the current time-sensitive flow delay is less than the acceptable upper limit of the worst delay, reduce the time-sensitive window value; When the current time-sensitive flow delay is equal to the acceptable upper limit of the worst delay, there is no need to adjust the time-sensitive window; When the current time-sensitive flow delay exceeds the acceptable upper limit of the worst delay, increase the time-sensitive window.

8. A multi-switch TSN dynamic scheduling system based on fuzzy control, characterized in that, The system comprises a processing unit configured to perform the steps of the fuzzy control-based multi-switch TSN dynamic scheduling method according to any one of claims 1-7.

9. An electronic device, comprising: It comprises a memory and a processor, the memory stores a computer program, and the processor is used to execute the program in the memory to realize the fuzzy control-based multi-switch TSN dynamic scheduling method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The storage medium stores the multi-switch TSN dynamic scheduling method based on fuzzy control in any one of claims 1-7.

Images