A collaborative scheduling system for industrial IoT equipment based on edge computing

Abstract

The present invention relates to the field of data processing technology, and more specifically, to an edge computing-based collaborative scheduling system for industrial Internet of Things devices. The system comprises: a data acquisition module for acquiring the load, delay time, path length between nodes, and time window of each transmission moment for each node in a data stream transmission node graph; a volatility calculation module for calculating the initial volatility of the node at the transmission moment; a target volatility speed calculation module for calculating the target volatility speed of the node at the transmission moment; and a scheduling module for using the target volatility speed of the node at the transmission moment in an ant colony algorithm to control path selection, thereby achieving collaborative scheduling of industrial Internet of Things devices and effectively improving the accuracy and efficiency of collaborative scheduling of industrial Internet of Things devices.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology. More specifically, the present invention relates to an industrial Internet of Things equipment collaborative scheduling system based on edge computing. Background Art

[0002] Edge computing aims to shift computing and data processing from the cloud to edge devices closer to the data source. With the continuous development of IoT technology, edge computing has become a key means of realizing the Industrial Internet of Things (IIoT). Edge computing and cloud computing complement each other in the IIoT. Edge computing, located close to the data source, provides low-latency, real-time processing capabilities, while cloud computing, with its powerful computing and storage capabilities, is suitable for processing large amounts of data. By synergizing these two approaches, efficient task offloading and optimal resource utilization can be achieved.

[0003] The ant colony algorithm (ACO) is a heuristic algorithm that simulates the foraging behavior of ants. By simulating the pheromone release of ants during their search for food, it gradually optimizes the solution. In the Industrial Internet of Things (IIoT) environment, data typically passes through multiple routers from the collection node to the termination node of the system platform. In edge computing environments, the ACO can optimize routing at the edge gateway, shortening data transmission paths and avoiding congested nodes, thereby reducing network latency and improving the coordinated scheduling efficiency of IIoT devices.

[0004] However, the pheromone evaporation rate in the ant colony algorithm is fixed. In complex or dynamically changing environments, a too fast evaporation rate can cause the algorithm to converge prematurely and miss the global optimal solution. A too slow evaporation rate can lead to slow convergence and low efficiency, ultimately affecting the accuracy of collaborative scheduling of IIoT devices. Therefore, accurately adjusting the pheromone evaporation rate of the ant colony algorithm to achieve the optimal scheduling results for IIoT devices is a pressing issue. Summary of the Invention

[0005] To solve the technical problem of how to accurately adjust the pheromone volatilization speed when the ant colony algorithm obtains the optimal scheduling results for industrial Internet of Things devices, the present invention proposes an industrial Internet of Things device collaborative scheduling system based on edge computing. The system includes the following modules:

[0006] The data acquisition module is used to obtain the load, delay time, and path length between nodes at each transmission moment of each node in the data flow transmission node diagram; obtain the time window of each node transmission moment; the volatility calculation module is used to calculate the volatility adjustment factor of the node at the transmission moment based on the maximum load, load variance, and maximum delay time of the node in the time window at the transmission moment; adjust the preset baseline volatility rate according to the volatility adjustment factor of the node at the transmission moment to obtain the initial volatility rate of the node at the transmission moment; the target volatility speed calculation module is used to calculate the target volatility speed of the node at the transmission moment:

[0007] ;

[0008] For the The target volatility of a node at the i-th transmission time is, 、 are the maximum and minimum values ​​of the data flow transmission path respectively, For the first The length of the data flow transmission path of each node, For the The initial volatility rate of each node at the i-th transmission moment is determined by the node volatility factor; and the scheduling module is used to control the path selection using the target volatility rate of the node at the transmission moment in the ant colony algorithm to achieve collaborative scheduling of industrial Internet of Things devices.

[0009] The present invention can effectively improve the efficiency and accuracy of the collaborative scheduling of industrial Internet of Things devices by dynamically adjusting the ant colony algorithm to achieve the pheromone volatilization rate in the process of selecting industrial Internet of Things device nodes. In the process of dynamic adjustment, the present invention determines the volatilization adjustment factor by obtaining the load and delay time of the device node, which can accurately achieve dynamic load balancing and increase the pheromone volatilization rate on high-load or high-latency nodes, so as to accurately adjust the traffic to low-load or low-latency paths, avoid network congestion, and effectively improve the collaborative scheduling efficiency of industrial Internet of Things devices. In addition, the present invention takes into account that the path length between nodes is a static feature of node selection. Therefore, when determining the target volatilization rate of the node, the present invention also corrects the initial volatilization rate by the path length, sets a smaller volatilization speed for the shorter path node, and can increase the probability of ants selecting the node path, thereby quickly converging to the optimal path, effectively improving the accuracy and efficiency of the collaborative scheduling of industrial Internet of Things devices.

[0010] According to an edge computing-based collaborative scheduling system for industrial Internet of Things devices provided by the present invention, the load, delay time, and path length between nodes of each node in the data flow transmission node diagram are obtained at each transmission moment, and the method also includes: taking the local terminal as the starting node, the cloud platform as the ending node, the router between the local terminal and the cloud platform as the intermediate node, and the direction from the starting node to the ending node as the transmission direction to construct a data flow transmission node diagram.

[0011] According to an edge computing-based collaborative scheduling system for industrial Internet of Things devices provided by the present invention, the load, delay time, and path length between nodes of each node in the data stream transmission node diagram at each transmission moment are obtained, including: obtaining the node load and delay time from the log of the node-related device at each transmission moment, and obtaining the path length between the nodes based on the number of hops between the nodes.

[0012] According to an edge computing-based collaborative scheduling system for industrial Internet of Things devices provided by the present invention, the time window for each transmission moment of the node is obtained, including: presetting the time window length of the node to m, obtaining m historical transmission moments from the historical transmission moments before the transmission moment to construct the time window of the transmission moment.

[0013] The present invention takes into account the dynamics, timing correlation, and statistical regularity of network load and delay, and their characteristics will be reflected in historical time periods. Therefore, by analyzing the load and delay time stability of each node at the historical transmission moments before the current transmission moment, the present invention can accurately obtain the difference between the load and delay time at the current collection moment and the historical data, thereby accurately obtaining the volatility adjustment factor.

[0014] According to an edge computing-based collaborative scheduling system for industrial Internet of Things devices provided by the present invention, the step of calculating the volatility adjustment factor of the node at the transmission moment includes:

[0015] ;

[0016] For the The volatility adjustment factor of a node at the i-th transmission time, 、 、 Respectively The maximum load, load variance, and delay time of each node in the time window of the i-th transmission time are 、 、 are the maximum load, maximum load variance, and maximum delay time of all nodes in the time window of the i-th transmission moment, respectively. is an exponential function with base e.

[0017] The present invention calculates the volatility adjustment factor by obtaining the load and delay time characteristics in the time window of the node's current transmission moment. When adjusting the pheromone volatility rate based on this, it can effectively reduce the attractiveness of high-load paths or high-delay paths, avoid the influx of more traffic causing delay deterioration, and reduce the possibility of congestion spreading, thereby breaking the local optimum and discovering a more balanced transmission path.

[0018] According to an edge computing-based collaborative scheduling system for industrial Internet of Things devices provided by the present invention, adjusting a preset baseline volatility rate according to a volatility adjustment factor of a node at a transmission time to obtain an initial volatility rate of the node at the transmission time includes:

[0019] ;

[0020] For the The initial volatility rate of a node at the i-th transmission time is, is the preset benchmark volatility rate, For the The volatility adjustment factor of a node at the i-th transmission time.

[0021] According to an edge computing-based collaborative scheduling system for industrial Internet of Things devices provided by the present invention, the target volatility speed of a node at the transmission moment is used in the ant colony algorithm to control path selection, including: obtaining the next adjacent node of the node in the node transmission direction; using the target volatility speed of the node at the transmission moment as the target volatility speed of the pheromone on the path between the node at the transmission moment and its next adjacent node; and calculating the path selection probability based on the target volatility speed of the pheromone on the path between the node and its next adjacent node at each transmission moment.

[0022] By using a node's target pheromone volatility rate at the moment of transmission as the target pheromone volatility rate along the path between that node and its next adjacent node at that moment, the present invention uses the static path length as a benchmark to avoid excessive volatility fluctuations when dynamic load or latency fluctuates dramatically due to burst traffic. The static path length prevents the algorithm from prematurely converging on a dynamically high-quality but non-shortest path, effectively improving path selection accuracy and algorithm efficiency.

[0023] According to an edge computing-based collaborative scheduling system for industrial Internet of Things devices provided by the present invention, the path selection probability is calculated based on the target pheromone volatility speed of the node on the path between the node and its next adjacent node at each transmission moment, including: substituting the target volatility speed of the node at the transmission moment into the iterative update formula to obtain the pheromone concentration on the path between the node at the transmission moment and its next adjacent node; and calculating the selection probability of the path between the node at the transmission moment and its next adjacent node based on the pheromone concentration, path length, preset pheromone importance factor and heuristic information importance factor on the path between the node at the transmission moment and its next adjacent node.

[0024] According to the present invention, a collaborative scheduling system for industrial Internet of Things devices based on edge computing is provided. The ant colony algorithm uses the target volatility speed of the node at the transmission time to control path selection to achieve collaborative scheduling of industrial Internet of Things devices, including: setting the number of ants, each ant randomly selecting a starting point, determining the ant's next node based on the selection probability of the path between the node and its next adjacent node at each transmission time, and after the ant arrives at the next node, updating the pheromone concentration of the path between the node and the next node in the data stream transmission node graph to achieve collaborative scheduling of industrial Internet of Things devices.

[0025] According to an edge computing-based collaborative scheduling system for industrial Internet of Things devices provided by the present invention, the collaborative scheduling of industrial Internet of Things devices is implemented, and further includes: determining a maintenance method for a node based on the selection probability of the node.

[0026] The present invention takes into account that the nodes with higher selection probabilities are more likely to be selected as the paths passed by ants, and the possibility of loss is also higher. Therefore, the present invention effectively improves the use time and use effect of device nodes by setting the corresponding maintenance mode of the nodes according to the selection probability of the nodes.

[0027] The present invention has the following technical effects:

[0028] Based on the above technical solution, the present invention can effectively improve the efficiency and accuracy of the collaborative scheduling of industrial Internet of Things devices by dynamically adjusting the ant colony algorithm to achieve the pheromone volatilization rate in the process of selecting industrial Internet of Things device nodes. In the process of dynamic adjustment, the present invention determines the volatilization adjustment factor by obtaining the load and delay time of the device node, which can accurately achieve dynamic load balancing, increase the pheromone volatilization rate on high-load or high-latency nodes, so as to accurately adjust the traffic to low-load or low-latency paths, avoid network congestion, and effectively improve the collaborative scheduling efficiency of industrial Internet of Things devices. In addition, the present invention takes into account that the path length between nodes is a static feature of node selection. Therefore, when determining the target volatilization rate of the node, the present invention also corrects the initial volatilization rate by the path length, sets a smaller volatilization speed for the shorter path node, and can increase the probability of ants selecting the node path, thereby quickly converging to the optimal path, effectively improving the accuracy and efficiency of the collaborative scheduling of industrial Internet of Things devices. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 This is a system block diagram of an industrial Internet of Things equipment collaborative scheduling system based on edge computing in an embodiment of the present invention. DETAILED DESCRIPTION

[0030] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, but not all of the embodiments.

[0031] The present invention provides an industrial Internet of Things equipment collaborative scheduling system based on edge computing, which can accurately realize the collaborative scheduling of industrial Internet of Things equipment through the ant colony algorithm, thereby effectively improving resource utilization.

[0032] like Figure 1 As shown, Figure 1 This is a system block diagram of an industrial Internet of Things device collaborative scheduling system based on edge computing in an embodiment of the present invention. The system includes modules 101 to 104, which are described in detail below.

[0033] The data acquisition module 101 is used to obtain the load, delay time, and path length between nodes of each node at each transmission moment in the data stream transmission node diagram, and obtain the time window of each node at each transmission moment.

[0034] For example, in an embodiment of the present invention, the local terminal can be used as the starting node, the cloud platform as the ending node, the router between the local terminal and the cloud platform as the intermediate node, and the direction from the starting node to the ending node as the transmission direction to construct a data flow transmission node graph.

[0035] Specifically, the transmission direction is obtained according to the data flow direction and the protocol.

[0036] It should be noted that in the data flow transmission node diagram, the data flow acquisition device is the starting node, and the cloud device that receives the data flow is the ending node. The data flow from the starting node to the ending node usually includes multiple nodes. For example, when the local device camera captures the video stream and uploads it to the cloud, there will be multiple routing device nodes to choose from. The connection method of these routing device nodes is usually a complex topology, that is, each node may not be connected to a unique number of other nodes at the same time. Task allocation and data routing between multiple devices require nanosecond-level coordinated scheduling, otherwise it may cause control command lag or production process interruption.

[0037] It's worth noting that the ant colony algorithm, by determining the path selection probabilities between nodes, can achieve coordinated scheduling of IIoT devices. In a data flow transmission node graph, real-time monitoring of each node's network load (such as bandwidth utilization and data flow) allows for dynamic avoidance of high-load areas, preventing data transmission congestion that can lead to delays, packet loss, and even system crashes. In the IIoT, excessive node latency can lead to delayed device operation, command loss, or misaligned execution.

[0038] Therefore, when obtaining the optimal pheromone volatilization speed between each node, the embodiment of the present invention can be determined by the node load size, load stability, and the size and stability of the delay time. When the node load is high and the load is relatively unstable, and the delay time is long and unstable, it is necessary to increase the pheromone volatilization rate, thereby reducing the probability of the node being selected by the data stream during the next transmission.

[0039] For example, in an embodiment of the present invention, obtaining the load and delay time of each node at each transmission moment in a data stream transmission node diagram includes: obtaining the load and delay time of the node from a log of a node-related device at each transmission moment.

[0040] The unit of the node load may be percentage, and the unit of the delay time may be milliseconds.

[0041] Specifically, when obtaining the delay time, a lower transmission frequency can be set due to the low computing power of the edge device.

[0042] For example, the transmission time interval may be set to one minute, and may be specifically set according to the size of the data stream, the load of related devices, and the delay time. The embodiment of the present invention does not impose any excessive restrictions on this.

[0043] It is understandable that network load and delay are dynamic, time-related and statistically regular. Therefore, the embodiment of the present invention can obtain its volatility adjustment factor by counting the load index and delay index of each node in the historical time period before the current transmission moment.

[0044] For example, in an embodiment of the present invention, obtaining the time window of each transmission moment of the node includes: presetting the time window length of the node to m, obtaining m historical transmission moments from the historical transmission moments before the transmission moment to construct the time window of the transmission moment.

[0045] Among them, m can be set to 5; the value of m can be set according to actual needs, and the embodiment of the present invention does not impose too many restrictions on this.

[0046] The volatility calculation module 102 is configured to calculate a volatility adjustment factor for a node at a transmission time based on the node's maximum load, load variance, and maximum delay time in a time window at the transmission time; and to adjust a preset baseline volatility based on the node's volatility adjustment factor at the transmission time to obtain the node's initial volatility at the transmission time.

[0047] The base volatility rate may be preset to 0.1; the base volatility rate may range from 0 to 1 and may be set according to actual needs.

[0048] It should be noted that the probability of an ant selecting a path is positively correlated with pheromone concentration: paths with higher pheromone concentrations are more likely to be selected by subsequent ants. If the load at a node's current transmission moment differs significantly from the load of the entire node in the data flow transmission node graph at the current transmission moment, and if the load fluctuates significantly, this indicates that the node is overloaded. Continuing to select this node will lead to packet backlogs, increased latency, and even network congestion and collapse. Correspondingly, if the node's latency is high (e.g., close to the network's maximum latency), continuing to select this node will result in extended packet queuing times and may even cause end-to-end latency to exceed the application's tolerance threshold.

[0049] Therefore, if the load and delay time values ​​of a node are both large and the stability is poor, it is necessary to increase the volatility rate to accelerate the attenuation of pheromones on the node path, reduce the probability of it being selected by subsequent ants, and thus guide traffic to paths with lower delay and load, avoiding a chain reaction of delay and overload.

[0050] For example, in an embodiment of the present invention, calculating the volatility adjustment factor of the node at the transmission time includes:

[0051] ;

[0052] For the The volatility adjustment factor of a node at the i-th transmission time, For the The maximum load of a node in the time window of the i-th transmission time is, For the The load variance of a node in the time window of the i-th transmission time is, For the The maximum delay time of a node in the time window of the i-th transmission time is, is the maximum load of all nodes in the time window of the i-th transmission time, is the maximum load variance of all nodes in the time window of the i-th transmission time, is the maximum delay time of all nodes in the time window of the i-th transmission time, is an exponential function with base e.

[0053] Among the above, It will Normalize, if The closer , explain the The closer the maximum load value of a node in the time window of the i-th transmission moment is to the global maximum load value, the higher the load intensity is, and the higher the possibility that the node has a high load is.

[0054] It will Normalize, if The closer , explain the The closer the maximum value of the load variance of a node in the time window of the i-th transmission moment is to the maximum value of the global node load variance, the higher the possibility that the node has high load fluctuation.

[0055] It will Normalize, if The closer , explain the The closer the maximum value of the delay time of a node in the time window of the i-th transmission time is to the global maximum value of the delay time, the higher the delay is, and the higher the possibility that the node has high delay.

[0056] In summary, The closer the maximum load of a node in the time window of the i-th transmission time is to the global maximum, the The closer the maximum load variance of a node in the time window of the i-th transmission moment is to the maximum global node load variance, and the closer the maximum delay time is to the maximum global delay time, the more likely it is that the node has high load and high delay at the current transmission moment. In order to reduce the probability of subsequent ants choosing this path, it is necessary to increase the volatilization speed of the pheromone concentration, that is, to increase the volatilization adjustment factor.

[0057] After the volatility adjustment factor of the node at the transmission moment is obtained based on the above method, the reference volatility rate can be adjusted according to the volatility adjustment factor of the node at each transmission moment, thereby accurately obtaining the initial volatility rate.

[0058] For example, in an embodiment of the present invention, adjusting a preset baseline volatility rate according to a volatility adjustment factor of a node at a transmission time to obtain an initial volatility rate of the node at the transmission time includes:

[0059] ;

[0060] For the The initial volatility rate of a node at the i-th transmission time is, is the preset benchmark volatility rate, For the The volatility adjustment factor of a node at the i-th transmission time.

[0061] In the above formula, The value of is less than or equal to 1, so the final The range of the initial volatility of a node at the i-th transmission time is: , ensuring that the path pheromone will not be quickly cleared due to an excessively high initial volatilization rate.

[0062] Based on the above steps, we can obtain the initial volatility rate of each node at each transmission moment. It is understandable that the node's load and delay time are dynamic, and so is the pheromone volatility rate. Therefore, the node's pheromone volatility rate can be dynamically adjusted based on the node's load and delay time. Furthermore, the node's path length remains static and is also a significant factor influencing the pheromone volatility rate. The lower the node's path length, the higher its path preference. In this case, its pheromone volatility rate needs to be reduced to provide a basis for subsequent ant selection.

[0063] Therefore, the evaporation speed of the path pheromone is not only related to the device load and delay time of the node, but also to the path length. The embodiment of the present invention continues to obtain the path length of the node through the following steps. By combining dynamic and static methods, the pheromone evaporation speed of the node at each transmission moment can be effectively improved, that is, continue to execute the following steps.

[0064] The target volatility calculation module 103 is configured to calculate the target volatility of the node at the transmission time according to the path length and initial volatility rate of the node.

[0065] For example, in an embodiment of the present invention, obtaining the path length between nodes in a data stream transmission node graph includes: obtaining the path length between the nodes according to the number of hops between the nodes.

[0066] If the node is a router, the path length between the nodes can be obtained according to the number of hops between the nodes through a routing protocol, which can be specifically set according to actual needs and will not be elaborated in detail in the embodiment of the present invention.

[0067] It is understandable that by transmitting the node graph through the data flow, each complete path from the starting node to the ending node can be obtained. If the data flow transmission path of the node is not unique, all the data flows passing through the The average value of the data flow transmission path of each node is taken as the The data flow transmission path length of each node. At nodes with shorter paths, the volatility of pheromones needs to be reduced to increase the probability of ants choosing the node, thereby accelerating convergence; while for nodes with longer paths, the volatility of pheromones needs to be enhanced to reduce the probability of ants choosing the node, thereby accelerating convergence. In addition, since the transmission direction of the data flow in the data flow transmission node diagram is from the starting node to the ending node, the path length of the node described in the embodiment of the present invention refers to the path length between the node and the next adjacent node in the transmission direction, while the data flow transmission path length is the complete data flow transmission path length from the starting node to the ending node.

[0068] For example, in an embodiment of the present invention, the target volatility of the node at the transmission time is calculated, and the specific relationship can be referred to as follows:

[0069] ;

[0070] For the The target volatility of a node at the i-th transmission time is, is the maximum value of the data flow transmission path, is the minimum value of the data flow transmission path, For the first The length of the data flow transmission path of each node, For the The initial volatility rate of a node at the i-th transmission time.

[0071] In the above formula, It is the extreme value of the transmission path, which is used as the benchmark value of different nodes to facilitate comparison between different nodes. Indicates the correction factor of the node path to the initial volatility, after the The closer the data flow transmission path length of each node is to the maximum value of the data flow transmission path, the closer the data flow transmission path length of each node is to the maximum value of the data flow transmission path. The longer the data flow transmission path length of each node, the lower the path length preference. In this case, it is necessary to increase the initial volatility rate according to the correction factor of the initial volatility rate to increase the pheromone volatility speed of the node and reduce the possibility of subsequent ants choosing this path.

[0072] Based on the above method, the target evaporation speed of each node at each transmission moment can be accurately obtained. The pheromone concentration after evaporation is updated based on its target evaporation speed, which can make ants tend to choose nodes with low load, low latency and short path length in each path node selection, effectively improving the collaborative scheduling effect and accuracy of ants in each node selection.

[0073] The scheduling module 104 is used to control path selection using the target volatility of the nodes at the transmission time in the ant colony algorithm to achieve collaborative scheduling of industrial Internet of Things devices.

[0074] It is understandable that in the collaborative scheduling of industrial IoT devices in edge computing, the ants in the ant colony algorithm represent data streams, and the ants’ selection of path nodes represents the data stream’s selection of path devices.

[0075] For example, in an embodiment of the present invention, the target volatility speed of a node at a transmission moment is used in an ant colony algorithm to control path selection, including: obtaining the next adjacent node of the node in the transmission direction of the node; using the target volatility speed of the node at the transmission moment as the target volatility speed of the pheromone on the path between the node at the transmission moment and its next adjacent node; and calculating the path selection probability based on the target volatility speed of the pheromone on the path between the node and its next adjacent node at each transmission moment.

[0076] For example, in an embodiment of the present invention, the path selection probability is calculated based on the target pheromone evaporation speed of the node on the path between the node and its next adjacent node at each transmission moment, including: substituting the target evaporation speed of the node at the transmission moment into the iterative update formula to obtain the pheromone concentration on the path between the node at the transmission moment and its next adjacent node; and calculating the selection probability of the path between the node at the transmission moment and its next adjacent node based on the pheromone concentration, path length, preset pheromone importance factor and heuristic information importance factor on the path between the node at the transmission moment and its next adjacent node.

[0077] Specifically, the iterative update formula is:

[0078] ;

[0079] In the above formula, For the The pheromone concentration on the path between a node and its next adjacent node at the i-th transmission time is, For the The initial volatility rate of a node at the i-th transmission time is, is the preset initial value of pheromone concentration.

[0080] The initial value of the pheromone concentration can be set to 13; the initial value of the pheromone concentration can be specifically set according to the number of nodes and the average path length.

[0081] Specifically, the number of ants can be set to 30. The pheromone importance factor controls the importance of pheromones in path selection and can be set to 1. The heuristic information importance factor controls the importance of the heuristic function in path selection and can be set to 1. The iteration termination condition can be preset as each ant completes the route selection or the iteration reaches a predetermined number of iterations, which can be set to 500.

[0082] For example, calculating the selection probability of the path between the node at the transmission moment and its next adjacent node based on the pheromone concentration, path length, preset pheromone importance factor, and heuristic information importance factor on the path between the node at the transmission moment and its next adjacent node includes:

[0083] ;

[0084] For the The probability of selecting a path between a node and its next adjacent node at the i-th transmission time is, For the The pheromone concentration on the path between a node and its next adjacent node at the i-th transmission time is: is the pheromone importance factor, is the heuristic information importance factor, For the The heuristic function of the path between a node and its next adjacent node at the i-th transmission time, For the The set of neighbor nodes of a node, For the The node index in the neighbor node set of the node, For the The node at the i-th transmission time and its neighbor node The pheromone concentration on the path between adjacent nodes, For the The node at the i-th transmission time and its neighbor node The heuristic function for the paths between adjacent nodes.

[0085] The heuristic function may be the inverse of the path length.

[0086] For example, in an embodiment of the present invention, the target volatility speed of the node at the transmission moment is used in the ant colony algorithm to control path selection to achieve collaborative scheduling of industrial Internet of Things devices, including: setting the number of ants, each ant randomly selecting a starting point, determining the next node of the ant based on the selection probability of the path between the node and its next adjacent node at each transmission moment, and after the ant arrives at the next node, updating the pheromone concentration of the path between the node and the next node in the data stream transmission node graph to achieve collaborative scheduling of industrial Internet of Things devices.

[0087] Specifically, the ants randomly sample the path between the node and its next adjacent node at each transmission moment, determine the next node to be reached, and update the pheromone concentration based on the pheromone concentration on the path between the node and its next adjacent node at the current transmission moment.

[0088] It is understandable that ants take into account the node load, delay time and path length in the process of determining the pheromone volatilization rate, and implement path selection based on this. This selection process is the process of collaborative scheduling.

[0089] For example, in an embodiment of the present invention, collaborative scheduling of industrial Internet of Things devices is implemented, and then the method further includes: determining a maintenance method of a node based on the selection probability of the node.

[0090] It is understandable that the higher the selection probability of a node, the greater the possibility that the node will be selected as the ant's forward route in random selection, that is, the greater the possibility that it will be used as a data flow transmission path. Long-term and high-frequency use of some nodes may easily lead to performance degradation or even failure due to overload. Therefore, when determining the maintenance method based on the node selection probability, a more complete maintenance method needs to be used for nodes with a higher selection probability.

[0091] For example, for nodes with a high probability of carrying main business, the port utilization can be monitored once an hour, and when it exceeds 70%, it will be automatically diverted to other top-of-rack switches.

[0092] The maintenance mode can be specifically set according to actual needs, and the embodiment of the present invention does not impose too many restrictions here.

[0093] It can be seen that in an embodiment of the present invention, when implementing collaborative scheduling of industrial interconnected devices based on edge computing, it can include a data acquisition module for obtaining the load, delay time, and path length between nodes of each node in the data stream transmission node diagram at each transmission moment; obtaining the time window of each node transmission moment; a volatility calculation module for calculating the volatility adjustment factor of the node at the transmission moment based on the maximum load, load variance, and maximum delay time of the node in the time window at the transmission moment; adjusting the preset baseline volatility rate according to the volatility adjustment factor of the node at the transmission moment to obtain the initial volatility rate of the node at the transmission moment; and a target volatility speed calculation module for calculating the target volatility speed of the node at the transmission moment:

[0094] ;

[0095] For the The target volatility of a node at the i-th transmission time is, 、 are the maximum and minimum values ​​of the data flow transmission path respectively, For the first The length of the data flow transmission path of each node, For the The initial volatility rate of each node at the i-th transmission moment; the scheduling module is used to use the target volatility rate of the node at the transmission moment to control the path selection in the ant colony algorithm to achieve collaborative scheduling of industrial Internet of Things devices, effectively improving the accuracy and efficiency of collaborative scheduling of industrial Internet of Things devices.

[0096] The above are all preferred embodiments of the present invention, and are not intended to limit the scope of protection of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the scope of protection of the present invention.

Claims

1. An industrial Internet of Things equipment collaborative scheduling system based on edge computing, characterized in that: Includes the following modules: The data acquisition module is used to obtain the load, delay time, and path length between nodes of each node at each transmission moment in the data flow transmission node diagram; and obtain the time window of each node at each transmission moment; a volatility calculation module configured to calculate a volatility adjustment factor for the node at the transmission moment based on the maximum load, load variance, and maximum delay time of the node in the time window at the transmission moment; and adjust a preset baseline volatility rate based on the volatility adjustment factor of the node at the transmission moment to obtain an initial volatility rate for the node at the transmission moment; The target volatility calculation module is used to calculate the target volatility of the node at the transmission time: ； For the The target volatility of a node at the i-th transmission time is, 、 are the maximum and minimum values ​​of the data flow transmission path respectively, For the first The length of the data flow transmission path of each node, For the The initial volatility rate of a node at the i-th transmission time; The scheduling module is used to control path selection using the target volatility of nodes at the transmission time in the ant colony algorithm to achieve collaborative scheduling of industrial IoT devices; The calculating the volatility adjustment factor of the node at the transmission moment includes: ； For the The volatility adjustment factor of a node at the i-th transmission time, 、 、 Respectively The maximum load, load variance, and delay time of each node in the time window of the i-th transmission time are 、 、 are the maximum load, maximum load variance, and maximum delay time of all nodes in the time window of the i-th transmission moment, respectively. is an exponential function with base e; The step of adjusting the preset reference volatility rate according to the volatility adjustment factor of the node at the transmission time to obtain the initial volatility rate of the node at the transmission time includes: ； For the The initial volatility rate of a node at the i-th transmission time is, is the preset benchmark volatility rate, For the The volatility adjustment factor of a node at the i-th transmission time; The method of using the target volatility of the node at the transmission time to control path selection in the ant colony algorithm includes: The next adjacent node of the node is obtained in the node transmission direction; the target evaporation speed of the node at the transmission moment is used as the target evaporation speed of the pheromone on the path between the node at the transmission moment and its next adjacent node; the path selection probability is calculated according to the target evaporation speed of the pheromone on the path between the node and its next adjacent node at each transmission moment.

2. The industrial Internet of Things equipment collaborative scheduling system based on edge computing according to claim 1 is characterized in that: The step of obtaining the load, delay time, and path length between nodes of each node in the data stream transmission node graph at each transmission moment also includes: The local terminal is taken as the starting node, the cloud platform is taken as the ending node, the router between the local terminal and the cloud platform is taken as the intermediate node, and the direction from the starting node to the ending node is taken as the transmission direction to construct a data flow transmission node graph.

3. The industrial Internet of Things equipment collaborative scheduling system based on edge computing according to claim 1 is characterized in that: The step of obtaining the load, delay time, and path length between nodes of each node in the data stream transmission node graph at each transmission moment includes: At each transmission moment, the node load and delay time are obtained from the logs of the node-related devices, and the path length between the nodes is obtained according to the number of hops between the nodes.

4. The industrial Internet of Things equipment collaborative scheduling system based on edge computing according to claim 1 is characterized in that: The acquiring of the time window of each transmission moment of the node includes: The time window length of the preset node is m, and m historical transmission moments are obtained from the historical transmission moments before the transmission moment to construct the time window of the transmission moment.

5. The industrial Internet of Things equipment collaborative scheduling system based on edge computing according to claim 1 is characterized in that: The calculating of the path selection probability according to the pheromone target volatilization speed on the path between the node and its next adjacent node at each transmission moment includes: Substitute the target evaporation speed of the node at the transmission time into the iterative update formula to obtain the pheromone concentration on the path between the node at the transmission time and its next adjacent node; Calculate the probability of selecting the path between the node at the transmission moment and its next adjacent node based on the pheromone concentration, path length, preset pheromone importance factor, and heuristic information importance factor on the path between the node at the transmission moment and its next adjacent node; The iterative update formula is: ； In the above formula, For the The pheromone concentration on the path between a node and its next adjacent node at the i-th transmission time is: For the The initial volatility rate of a node at the i-th transmission time is, is the preset initial value of pheromone concentration.

6. The industrial Internet of Things equipment collaborative scheduling system based on edge computing according to claim 1 is characterized in that: The method uses the target volatility of nodes at the transmission time in the ant colony algorithm to control path selection to achieve collaborative scheduling of industrial Internet of Things devices, including: The number of ants is set, and each ant randomly selects a starting point. The next node of the ant is determined based on the selection probability of the path between the node and its next adjacent node at each transmission moment. After the ant arrives at the next node, the pheromone concentration of the path between the node and the next node in the data flow transmission node graph is updated to realize the coordinated scheduling of industrial Internet of Things devices.

7. The industrial Internet of Things equipment collaborative scheduling system based on edge computing according to claim 6 is characterized in that: The implementation of collaborative scheduling of industrial IoT devices further includes: The node maintenance method is determined based on the node selection probability.

Images