Business topology automatic construction method and system based on industrial environment

Abstract

The invention discloses a business topology automation construction method and system based on an industrial environment, belongs to the field of industrial business topology automation, and aims to analyze a logical relationship between devices in real time based on a neural network, dynamically adjust node priorities and combine a topology optimization mechanism of a genetic algorithm / particle swarm optimization algorithm. The method reduces the false alarm rate of the system under the scenes of equipment firmware upgrade, configuration change and the like, adopts an event-triggered topology automatic update mechanism, eliminates the requirements of manual maintenance of a fingerprint library and a rule library, and realizes automatic discovery and construction of industrial service topology by monitoring the network state in real time and triggering adaptive reconstruction. A complete and accurate topological graph can be quickly generated without manual intervention, and a configuration backup and progressive switching mechanism is implemented in a topology reconstruction process to ensure service continuity.

Description

Technical Field

[0001] The present invention belongs to the field of industrial business topology automation, and in particular relates to a business topology automation construction method and system based on an industrial environment. Background Art

[0002] Nowadays, industrial control systems have become an important part of critical infrastructure. In industrial control systems, device identification and behavior monitoring usually rely on the following technical means: static fingerprint library comparison: asset identification is performed through static feature libraries such as preset device MAC addresses and firmware versions, which cannot adapt to dynamic changes in device behavior; threshold alarm mechanism: setting fixed thresholds can easily cause false positives or missed reports; single-dimensional monitoring: independent analysis based on network traffic or device logs is difficult to fully reflect the device status; rule engine driven: relying on manually written rule libraries to match known attack patterns, it is difficult to identify unknown threats and has high maintenance costs.

[0003] The above methods have poor dynamic adaptability when facing complex and changing industrial environments: they are unable to automatically identify scenarios such as firmware upgrades and configuration changes (the false alarm rate is as high as 35%); the feature dimension is single: it ignores the association between the timing characteristics of device behavior and the process context (the missed alarm rate exceeds 40%); the maintenance cost is high: the fingerprint library and detection rules need to be manually updated regularly (an average of 8 man-hours per week); and there is a serious detection lag: there is a 12-48 hour window between feature changes and the effectiveness of rules. Summary of the Invention

[0004] In view of the above-mentioned deficiencies in the prior art, the present application provides a method and system for automatically constructing a business topology based on an industrial environment.

[0005] In a first aspect, the present application proposes a method for automatically constructing a service topology based on an industrial environment, comprising the following steps:

[0006] Collect sensor data, control system data, and network communication data from industrial equipment, integrate and process the collected data, and build a real-time database of equipment nodes;

[0007] Based on the neural network, the logical relationship between the devices in the real-time database of the device nodes is identified, and the node priorities are dynamically sorted, wherein each device is a node, each node has a corresponding weight value, and the priority is dynamically sorted according to the size of the weight value to obtain the sorting result;

[0008] A genetic algorithm or a particle swarm optimization algorithm is used in combination with the sorting result to generate a first network topology structure, and a topology reconstruction strategy is dynamically called according to changes in service type and network status to perform a topology transformation on the first network topology structure to obtain a second network topology structure including a ring topology structure and a star topology structure, thereby achieving adaptive optimization of the topology structure;

[0009] Select the first or second network topology structure according to actual network requirements, and construct it into an industrial business topology graph by combining it with a graph theory algorithm;

[0010] By combining regular polling and event triggering, the operating status and network connection of the equipment are monitored in real time. When a new device is detected to be connected to the network, the topology update process is automatically triggered to collect the data of the new device and add it to the industrial business topology map to achieve topology update.

[0011] In some embodiments, using a genetic algorithm in combination with the sorting result to generate a first network topology structure includes:

[0012] Step A1: Randomly generate N binary matrices according to the encoding method to form an initial population, where each binary matrix represents an industrial network topology;

[0013] Step A2: Calculate the fitness value of each individual in the population according to the designed fitness function, where each individual represents a possible topological structure;

[0014] Step A3: Use roulette wheel selection or tournament selection to select individuals with higher fitness from the current population to enter the next generation population;

[0015] Step A4: Perform crossover operation on the selected individuals according to the set crossover probability;

[0016] Step A5: Mutate the individual's genes based on the mutation probability;

[0017] Step A6: Repeat steps A2-A5 until the set number of iterations is reached or the stopping condition is met, at which point the topology corresponding to the individual with the highest fitness in the population is the first network topology.

[0018] In some embodiments, the encoding method of the genetic algorithm includes: using binary encoding to represent the industrial network topology. Assume that there are n nodes in the industrial network, then an n×n binary matrix T can be constructed, and the elements T in the matrix are ij , when T ij =1, indicating that there is a connection between node i and node j; when T ij =0, it means there is no connection between node i and node j.

[0019] In some embodiments, a particle swarm optimization algorithm is used in combination with the sorting result to generate a second network topology structure, including:

[0020] Step B1: Randomly generate M n×n real number matrices as the initial position of the particle, and initialize the velocity matrix V of the particle. The velocity matrix elements Vij The value range is determined according to the actual situation, generally ;

[0021] Step B2: Calculate the fitness value of each particle according to the designed fitness function;

[0022] Step B3: Compare the fitness value of each particle's current position with the fitness value of its own historical optimal position. If the current position is better, update the individual optimal position; compare the fitness values ​​of all particles, find the particle with the highest fitness, and set its position as the global optimal position;

[0023] Step B4: Update the particle's velocity and position according to the following formula:

[0024] ; Where t represents the current number of iterations, r1 ij and r2 ij is a random number in the interval [0,1], P ij represents the individual historical optimal position of particle i, G ij Represents the global optimal position, constrains the updated position, and ensures that the matrix elements are in the range [0,1];

[0025] Step B5: Repeat steps B2-B4 until the maximum number of iterations is reached or the stopping condition is met, at which point the topology structure corresponding to the global optimal position is the first network topology structure.

[0026] In some embodiments, the particle representation of the particle swarm optimization algorithm includes: representing each particle as an n×n real number matrix X, where n is the number of industrial network nodes, and the matrix element X is ij Indicates the possibility or weight of the connection between node i and node j, with a value range of [0,1]. ij When it is close to 1, it means that the possibility of connection between node i and node j is high; when X ij When it is close to 0, it means that the connection possibility is small.

[0027] In some embodiments, the fitness function design of the genetic algorithm or particle swarm optimization algorithm includes: assuming that R represents a network reliability index, T represents a real-time index, S represents a scalability index, and C represents a cost-effectiveness index, and the value range of each index is normalized so that it is within the interval [0, 1]. Then, the fitness function formula is: Among them, w1, w2, w3, and w4 are weight coefficients, and w1+w2+w3+w4=1. The setting of weight coefficients needs to be adjusted according to the actual needs of the industrial network.

[0028] In some embodiments, the dynamic invocation of a topology reconstruction strategy based on changes in business type and network status includes: the conversion based on network status comprises: real-time monitoring of status parameters of the industrial network, and when the bandwidth utilization of a certain link in the network continuously exceeds a threshold, and the delay and packet loss rate significantly increase, determining that the current topology structure cannot meet network requirements, and converting the first network topology structure into a corresponding topology structure according to a preset conversion strategy;

[0029] During the conversion process, first back up the current network configuration and data, then gradually adjust the connection relationship of network devices, and perform network testing at the same time to ensure the continuity and stability of network services during the conversion process.

[0030] In some embodiments, performing topology conversion on the first network topology to obtain a second network topology including a ring topology and a star topology to achieve adaptive optimization of the topology includes:

[0031] The ring topology is generated as follows: in the binary encoding of the genetic algorithm, a continuous sequence of nodes is mandatory, so that these nodes are connected in sequence to form a ring, and each node is connected only to adjacent nodes on the ring, wherein the connection element corresponding to the central node and other nodes in the matrix is ​​1, and the other elements are 0. In the particle swarm optimization algorithm, the initial position and update rules of the particles are set so that the connections between the nodes tend to form a ring structure, and when updating the particle position, the probability of increasing the connection weight of the adjacent nodes is increased;

[0032] The star topology is generated by specifying a central node in the binary matrix encoding of the genetic algorithm so that all other nodes are connected only to the central node, wherein the connection elements corresponding to the central node and other nodes in the matrix are 1, and the other elements are 0. In the particle swarm optimization algorithm, when initializing particles, the connection weights of most nodes are concentrated on a specific node to form the prototype of the star structure, and the stability of this connection pattern is maintained in subsequent iterations.

[0033] The tree topology is: defining the connection relationship between the parent node and the child node through hierarchical coding;

[0034] The mesh topology is as follows: connection constraints are relaxed, connections are allowed to be established between more nodes, and the indicators in the fitness function are used to guide the formation of a mesh structure that needs to be converted.

[0035] In the second aspect, the present application proposes a business topology automation construction system based on an industrial environment, including a data acquisition module, a node configuration module, an industrial network topology structure analysis module, an industrial business topology map construction module and an automatic response module;

[0036] The data acquisition module is used to collect sensor data, control system data and network communication data of industrial equipment, integrate and process the collected data, and build a real-time database of equipment nodes;

[0037] The node configuration module is used to identify the logical relationship between devices in the device node real-time database based on a neural network, and dynamically sort the node priorities, wherein each device is a node, each node has a corresponding weight value, and dynamic priority sorting is performed according to the size of the weight value to obtain a sorting result;

[0038] The industrial network topology analysis module is configured to generate a first network topology using a genetic algorithm or a particle swarm optimization algorithm in combination with the sorting result, dynamically invoke a topology reconstruction strategy based on changes in service type and network status, and perform a topology transformation on the first network topology to obtain a second network topology including a ring topology and a star topology, thereby achieving adaptive optimization of the topology;

[0039] The industrial service topology map construction module is used to select the first or second network topology structure according to actual network requirements and construct it into an industrial service topology map by combining a graph theory algorithm;

[0040] The automatic response module is used to monitor the operating status and network connection status of the equipment in real time by combining regular polling and event triggering. When a new device is detected to be connected to the network, the topology update process is automatically triggered to collect the data of the new device and add it to the industrial business topology map to achieve topology update.

[0041] In a third aspect, the present application proposes an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above method when executing the computer program.

[0042] In a fourth aspect, the present application proposes a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above method are implemented.

[0043] Beneficial effects of the present invention:

[0044] Based on real-time analysis of the logical relationship between devices by neural networks and dynamic adjustment of node priorities, combined with the topology optimization mechanism of genetic algorithm / particle swarm optimization algorithm, the system reduces the false alarm rate in scenarios such as device firmware upgrades and configuration changes. It adopts an event-triggered topology automatic update mechanism to eliminate the need for manual maintenance of fingerprint libraries and rule libraries. By monitoring the network status in real time and triggering adaptive reconstruction, it realizes the automatic discovery and construction of industrial business topology. It can quickly generate a complete and accurate topology map without human intervention. During the topology reconstruction process, it implements configuration backup and progressive switching mechanism to ensure business continuity. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is the overall flow chart of the present invention.

[0046] Figure 2 This is a system principle block diagram of the present invention. DETAILED DESCRIPTION

[0047] The following will describe exemplary embodiments of the present invention in more detail with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments described herein; rather, these embodiments are provided to enable a more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0048] In the first aspect, this application proposes a method for automatically constructing a business topology based on an industrial environment, such as Figure 1 As shown, the following steps are included:

[0049] S100: Collects sensor data, control system data, and network communication data from industrial equipment, integrates and processes the collected data, and builds a real-time database of device nodes;

[0050] Continuously capture operational status data through multiple types of sensors (temperature, pressure, position) installed on industrial equipment (CNC machine tools, robots, conveyor belts, etc.); extract device information (device model, performance parameters, job tasks) from the control system using industrial communication protocols (such as OPC UA, Modbus TCP, and Profinet); automatically identify network connections between devices using network detection technologies (such as active ping, port scanning, and ARP snooping) to obtain IP addresses, open port numbers, and communication protocols; and fuse multi-source data to build a real-time database of device nodes for subsequent topology modeling.

[0051] S200: Identifying logical relationships between devices in the device node real-time database based on a neural network, dynamically sorting node priorities, wherein each device is a node, each node has a corresponding weight value, and dynamically sorting priorities according to the weight values ​​to obtain a sorting result;

[0052] Based on the data in the real-time database of device nodes, each device is abstracted into a graph node. A graph structure is constructed based on the collected multi-dimensional features (device type, communication frequency, data flow, and criticality level). Graph neural networks (GCN) or long short-term memory networks (LSTM) are used to analyze the collaborative dependencies between devices, determine the communication strength between devices, and obtain a ranking result. Alternatively, after the sensor data is collected and uploaded to the platform, the platform is used to dynamically associate devices with the configured data sets. Each graph node has a corresponding weight value, and dynamic priority sorting is performed according to the size of the weight value. The communication strength between devices is determined based on the priority to obtain the ranking result.

[0053] S300: Using a genetic algorithm or a particle swarm optimization algorithm in combination with the sorting result to generate a first network topology structure, dynamically calling a topology reconstruction strategy based on service types and network status changes, performing a topology transformation on the first network topology structure to obtain a second network topology structure including a ring topology structure and a star topology structure, thereby achieving adaptive optimization of the topology structure;

[0054] In some embodiments, using a genetic algorithm in combination with the sorting result to generate a first network topology structure includes:

[0055] Step A1: Randomly generate N binary matrices according to the encoding method to form an initial population, where each binary matrix represents an industrial network topology;

[0056] Step A2: Calculate the fitness value of each individual in the population according to the designed fitness function, where each individual represents a possible topological structure;

[0057] Step A3: Use roulette wheel selection or tournament selection to select individuals with higher fitness from the current population to enter the next generation population;

[0058] Step A4: Perform crossover operation on the selected individuals according to the set crossover probability;

[0059] Step A5: Mutate the individual's genes based on the mutation probability;

[0060] Step A6: Repeat steps A2-A5 until the set number of iterations is reached or the stopping condition is met, at which point the topology corresponding to the individual with the highest fitness in the population is the first network topology.

[0061] In some embodiments, the encoding method of the genetic algorithm includes: using binary encoding to represent the industrial network topology. Assume that there are n nodes in the industrial network, then an n×n binary matrix T can be constructed, and the elements T in the matrix are ij , when T ij =1, indicating that there is a connection between node i and node j; when T ij =0, it means there is no connection between node i and node j.

[0062] Among them, the parameters of the genetic algorithm are set as:

[0063] 1. Population size: This is typically set to 50-200. A larger population size increases population diversity and prevents the algorithm from prematurely falling into local optimal solutions, but it increases computational complexity and time. A smaller population size is computationally efficient but may not fully search the solution space. In industrial network topology optimization scenarios, if the network is large and complex, a larger population size, such as 150, can be selected; if the network is small, a population size of around 80 can be selected.

[0064] 2. Number of iterations: Generally set between 100 and 500. If the number of iterations is too few, the algorithm may not be able to find a better solution; if the number of iterations is too high, computing resources and time will be wasted. For simple industrial network topology optimization, 200 iterations may be sufficient to find a better solution; for complex networks, 300-400 iterations may be required. Crossover probability: The value range is usually between 0.6 and 0.9. When the crossover probability is high, it can speed up the algorithm's search speed and promote the propagation and combination of excellent genes, but too high a crossover probability may destroy excellent individuals; when the crossover probability is low, the algorithm converges more slowly. In practical applications, the crossover probability can be set to 0.8 first and then adjusted according to the algorithm's performance.

[0065] 3. Mutation Probability: Generally, this value is between 0.001 and 0.01. Mutation increases population diversity and prevents the algorithm from falling into local optima. A high mutation probability will make the algorithm approach random search, while a low one will not effectively avoid local optima. You can initially set the mutation probability to 0.005 to observe the optimization effect of the algorithm.

[0066] In some embodiments, a particle swarm optimization algorithm is used in combination with the sorting result to generate a second network topology structure, including:

[0067] Step B1: Randomly generate M n×n real number matrices as the initial position of the particle, and initialize the velocity matrix V of the particle. The velocity matrix elements V ij The value range of is determined according to the actual situation, generally [-1,1];

[0068] Step B2: Calculate the fitness value of each particle according to the designed fitness function;

[0069] Step B3: Compare the fitness value of each particle's current position with the fitness value of its own historical optimal position. If the current position is better, update the individual optimal position; compare the fitness values ​​of all particles, find the particle with the highest fitness, and set its position as the global optimal position;

[0070] Step B4: Update the particle's velocity and position according to the following formula:

[0071] ;

[0072] Among them, t represents the current iteration number, r1 ij and r2 ij is a random number in the interval [0,1], P ij represents the individual historical optimal position of particle i, G ij Represents the global optimal position, constrains the updated position, and ensures that the matrix elements are in the range [0,1];

[0073] Step B5: Repeat steps B2-B4 until the maximum number of iterations is reached or the stopping condition is met, at which point the topology structure corresponding to the global optimal position is the first network topology structure.

[0074] In some embodiments, the particle representation of the particle swarm optimization algorithm includes: representing each particle as an n×n real number matrix X, where n is the number of industrial network nodes, and the matrix element X is ij Indicates the possibility or weight of the connection between node i and node j, with a value range of [0,1]. ij When it is close to 1, it means that the possibility of connection between node i and node j is high; when X ij When it is close to 0, it means that the connection possibility is small.

[0075] Among them, the parameters of the particle swarm optimization algorithm are set as:

[0076] 1. Particle swarm size: Typically set between 30 and 100 particles. Similar to genetic algorithms, a larger swarm size increases search power but also increases computational complexity. For medium-sized industrial networks, a swarm size of 50 particles is recommended.

[0077] 2. Maximum number of iterations: Generally 100-300. This number should be selected based on the complexity of the network and the availability of computing resources. Complex networks can have the number of iterations increased appropriately. 3. Learning factors c1 and c2: These values ​​are typically between [1, 2]. c1 and c2 represent the degree to which a particle learns from its own historical optimal position and the group's historical optimal position, respectively. Generally, c1 = c2 = 1.5, allowing particles to fully utilize both their own experience and the experience of the group. Inertia weight ω: This value ranges from [0.4, 0.9]. A larger inertia weight favors global search, while a smaller inertia weight favors local search. In the early stages of the algorithm, a larger ω value, such as 0.8, can be set to speed up the global search. In the later stages of the algorithm, the ω value can be gradually reduced, such as to 0.5, for a more refined local search.

[0078] In some embodiments, the fitness function design of the genetic algorithm or particle swarm optimization algorithm includes: assuming that R represents a network reliability index, T represents a real-time index, S represents a scalability index, and C represents a cost-effectiveness index, and the value range of each index is normalized so that it is within the interval [0, 1]. Then, the fitness function formula is: Among them, w1, w2, w3, and w4 are weight coefficients, and w1+w2+w3+w4=1. The setting of weight coefficients needs to be adjusted according to the actual needs of the industrial network.

[0079] In some embodiments, the dynamic invocation of a topology reconstruction strategy based on changes in business type and network status includes: the conversion based on network status comprises: real-time monitoring of status parameters of the industrial network, and when the bandwidth utilization of a certain link in the network continuously exceeds a threshold, and the delay and packet loss rate significantly increase, determining that the current topology structure cannot meet network requirements, and converting the first network topology structure into a corresponding topology structure according to a preset conversion strategy;

[0080] During the conversion process, first back up the current network configuration and data, then gradually adjust the connection relationship of network devices, and perform network testing at the same time to ensure the continuity and stability of network services during the conversion process.

[0081] In some embodiments, performing topology conversion on the first network topology to obtain a second network topology including a ring topology and a star topology to achieve adaptive optimization of the topology includes:

[0082] The ring topology is generated as follows: in the binary encoding of the genetic algorithm, a continuous sequence of nodes is mandatory, so that these nodes are connected in sequence to form a ring, and each node is connected only to adjacent nodes on the ring, wherein the connection element corresponding to the central node and other nodes in the matrix is ​​1, and the other elements are 0. In the particle swarm optimization algorithm, the initial position and update rules of the particles are set so that the connections between the nodes tend to form a ring structure, and when updating the particle position, the probability of increasing the connection weight of the adjacent nodes is increased;

[0083] The star topology is generated by specifying a central node in the binary matrix encoding of the genetic algorithm so that all other nodes are connected only to the central node, wherein the connection elements corresponding to the central node and other nodes in the matrix are 1, and the other elements are 0. In the particle swarm optimization algorithm, when initializing particles, the connection weights of most nodes are concentrated on a specific node to form the prototype of the star structure, and the stability of this connection pattern is maintained in subsequent iterations.

[0084] The tree topology is: defining the connection relationship between the parent node and the child node through hierarchical coding;

[0085] The mesh topology is as follows: connection constraints are relaxed, connections are allowed to be established between more nodes, and the indicators in the fitness function are used to guide the formation of a mesh structure that needs to be converted.

[0086] Generation of polymorphic topology:

[0087] 1. Ring Topology Generation: In genetic algorithms or particle swarm optimization (PSO), ring topology generation is achieved by constraining the encoding or particle representation. For example, in the binary encoding of a genetic algorithm, a continuous sequence of nodes is mandatory, so that these nodes are connected in sequence to form a ring, and each node is connected only to adjacent nodes in the ring (i.e., the corresponding position element in the matrix is ​​1, and all other non-adjacent positions are 0). In PSO, the initial positions of the particles and the update rules are set so that the connections between nodes tend to form a ring structure. For example, when updating the particle position, the probability of increasing the connection weight of adjacent nodes is increased.

[0088] 2. Star Topology Generation: Constraints are also added to the algorithm. In the genetic algorithm's binary matrix encoding, a central node is designated, and all other nodes are connected only to this central node (elements in the matrix corresponding to connections between the central node and other nodes are 1, and all other elements are 0). In the particle swarm optimization algorithm, when initializing particles, the connection weights of most nodes are concentrated on a specific node, forming the prototype of a star structure. This connection pattern is then maintained in a stable manner during subsequent iterations.

[0089] 3. Generation of other topological structures: For structures such as tree topology and mesh topology, they can be achieved by designing more complex constraints and coding rules.

[0090] For example, tree topology can define the connection relationship between parent nodes and child nodes through hierarchical coding; mesh topology relaxes connection constraints and allows more nodes to establish connections, but can use indicators such as reliability in the fitness function to guide the formation of a reasonable mesh structure.

[0091] Conversion between topologies:

[0092] Network status-based conversion: Real-time monitoring of industrial network status parameters such as bandwidth utilization, latency, and packet loss rate. When the bandwidth utilization of a certain link in the network consistently exceeds a threshold (e.g., 80%), and latency and packet loss rate increase significantly, the current topology is determined to be unable to meet network requirements. At this point, the system converts the current topology to a more suitable one based on a pre-set conversion strategy. For example, if the current star topology suffers from network performance degradation due to excessive load on the central node, it can be converted to a ring or distributed mesh topology. This can be achieved by adding redundant paths to disperse traffic and improve network reliability and performance.

[0093] Conversion based on actual needs: Topology conversion is performed when actual industrial network requirements change. For example, adding a new production line in a factory places higher demands on network scalability and real-time performance. If the current simple bus topology cannot meet the access and data transmission needs of the newly added devices, an algorithm can be used to re-optimize and generate a tree topology or hybrid topology to adapt to the new requirements. During the conversion process, the current network configuration and data are first backed up, and then the connection relationships between network devices are gradually adjusted. Network testing is also performed to ensure network service continuity and stability during the conversion process.

[0094] S400: Select the first or second network topology structure according to actual network requirements, and construct it into an industrial service topology graph by combining it with a graph theory algorithm;

[0095] The system automatically matches topology constraints based on the type of service carried by the industrial network (such as real-time control, data acquisition, and remote debugging). For example, in real-time control scenarios, the system prioritizes low latency and high reliability, adopting a star topology. In remote debugging scenarios, where fault tolerance is a priority, the system recommends a mesh topology.

[0096] A graph theory algorithm (Minimum Spanning Tree (MST)) is used to generate a topology that meets specific business requirements. The goals of the MST algorithm are:

[0097]

[0098] in, The communication cost between nodes u and v is solved by Prim's algorithm or Kruskal's algorithm to obtain the optimal MST structure, and then presented to the operation and maintenance personnel in a graphical form.

[0099] In addition, the system provides topology switching strategies based on historical optimization experience and current network status. For example, if the system detects that the packet loss rate of a link exceeds a threshold of 10, it will switch from a star topology to a partial mesh topology to enhance fault tolerance.

[0100] S500: By combining regular polling and event triggering, the operating status and network connection status of the equipment are monitored in real time. When a new device is detected to be connected to the network, the topology update process is automatically triggered to collect the data of the new device and add it to the industrial business topology map to achieve topology update.

[0101] 1. Polling Monitoring Mechanism Design

[0102] 1. Hierarchical polling cycle strategy

[0103] The three-level polling system is divided according to the importance of equipment and business sensitivity as shown in Table 1:

[0104] Table 1

[0105]

[0106] Dynamic adjustment formula: Where: T n : adjusted period, T0: initial period, k: adjustment coefficient (k=0.5 for core equipment, k=1 for common equipment), C load : Current system load, C th : load threshold;

[0107] 2. Polling Protocol and Data Collection Method

[0108] Industrial equipment: OPC UA protocol: 1s cycle, collects real-time parameters such as motor speed and temperature; Modbus RTU: 500ms cycle, polls PLC register status; Network equipment: SNMP v3: 30s cycle, obtains interface traffic (IF-MIB) and CPU utilization; ICMP ping: 2s cycle, monitors network connectivity (packet loss rate, latency);

[0109] Data cache strategy: Use a ring buffer to store the data of the latest 100 polling cycles.

[0110] 2. Event-triggered monitoring mechanism

[0111] 1. Design of device status trigger conditions:

[0112] Threshold type trigger (taking CNC machine tools as an example), as shown in Table 2:

[0113] Table 2

[0114]

[0115] State transition trigger:

[0116] Define a state machine model (e.g., Run → Standby → Fault) that is triggered when the following state changes occur:

[0117] Abnormal transition: running → fault (no normal shutdown process)

[0118] Timeout transition: Standby → Running timeout (exceeding the expected startup time by 30s)

[0119] 2. Network connection trigger conditions

[0120] Link layer triggers: Sudden packet loss: The packet loss rate increases from <1% to >10% for three consecutive ICMP packets; Bandwidth drop: The interface bandwidth utilization drops suddenly from 70% to 20% (possibly a link outage);

[0121] Application layer trigger: Service response timeout: HTTP request does not respond for more than 500ms (triggering the reconnection mechanism); abnormal session termination: TCP connection RST packet frequency > 5 times / minute).

[0122] In the second aspect, this application proposes a business topology automation construction system based on industrial environment, such as Figure 2 As shown, it includes data acquisition module, node configuration module, industrial network topology analysis module, industrial business topology map construction module and automatic response module;

[0123] The data acquisition module is used to collect sensor data, control system data and network communication data of industrial equipment, integrate and process the collected data, and build a real-time database of equipment nodes;

[0124] The node configuration module is used to identify the logical relationship between devices in the device node real-time database based on a neural network, and dynamically sort the node priorities, wherein each device is a node, each node has a corresponding weight value, and dynamic priority sorting is performed according to the size of the weight value to obtain a sorting result;

[0125] The industrial network topology analysis module is configured to generate a first network topology using a genetic algorithm or a particle swarm optimization algorithm in combination with the sorting result, dynamically invoke a topology reconstruction strategy based on changes in service type and network status, and perform a topology transformation on the first network topology to obtain a second network topology including a ring topology and a star topology, thereby achieving adaptive optimization of the topology;

[0126] The industrial service topology map construction module is used to select the first or second network topology structure according to actual network requirements and construct it into an industrial service topology map by combining a graph theory algorithm;

[0127] The automatic response module is used to monitor the operating status and network connection status of the equipment in real time by combining regular polling and event triggering. When a new device is detected to be connected to the network, the topology update process is automatically triggered to collect the data of the new device and add it to the industrial business topology map to achieve topology update.

[0128] In a third aspect, the present application proposes an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above method when executing the computer program.

[0129] In a fourth aspect, the present application proposes a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above method are implemented.

[0130] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0131] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant description of other embodiments.

[0132] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this disclosure.

[0133] In the embodiments provided in the present disclosure, it should be understood that the disclosed apparatus / computer equipment and methods can be implemented in other ways. For example, the apparatus / computer equipment embodiments described above are merely schematic. For example, the division of modules or units is merely a logical function division. In actual implementation, there may be other division methods. Multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces, indirect coupling or communication connection of the apparatus or unit, which may be electrical, mechanical or other forms.

[0134] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0135] In addition, the functional units in the various embodiments of the present disclosure may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0136] If the integrated module / unit is implemented as a software functional unit and sold or used as a standalone product, it can be stored in a computer-readable storage medium. Based on this understanding, the present disclosure can implement all or part of the process steps in the above-mentioned method embodiments by using a computer program to instruct the relevant hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, the computer program can implement the steps of each of the above-mentioned method embodiments. The computer program can include computer program code, which can be in source code form, object code form, executable file, or some intermediate form. Computer-readable media can include: any entity or device capable of carrying computer program code, recording media, USB flash drives, removable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunications signals, and software distribution media. It should be noted that the content included in computer-readable media can be appropriately increased or decreased based on the requirements of legislation and patent practice in a jurisdiction. For example, in some jurisdictions, based on legislation and patent practice, computer-readable media does not include electrical carrier signals and telecommunications signals.

[0137] The above are only preferred embodiments of the present invention. It should be pointed out that various modifications and improvements made by those skilled in the art without departing from the present technical solution should also be deemed to fall within the scope of protection required by this solution.

Claims

1. A method for automatically constructing a service topology based on an industrial environment, characterized by: The following steps are involved: Collect sensor data, control system data, and network communication data from industrial equipment, integrate and process the collected data, and build a real-time database of equipment nodes; Based on the neural network, the logical relationship between the devices in the real-time database of the device nodes is identified, and the node priorities are dynamically sorted, wherein each device is a node, each node has a corresponding weight value, and the priority is dynamically sorted according to the size of the weight value to obtain the sorting result; A genetic algorithm or a particle swarm optimization algorithm is used in combination with the sorting result to generate a first network topology structure, and a topology reconstruction strategy is dynamically called according to changes in service type and network status to perform a topology transformation on the first network topology structure to obtain a second network topology structure including a ring topology structure and a star topology structure, thereby achieving adaptive optimization of the topology structure; Select the first or second network topology structure according to actual network requirements, and construct it into an industrial business topology graph by combining it with a graph theory algorithm; By combining regular polling and event triggering, the operating status and network connection of the equipment are monitored in real time. When a new device is detected to be connected to the network, the topology update process is automatically triggered to collect the data of the new device and add it to the industrial business topology map to achieve topology update.

2. The method according to claim 1, wherein: Generating a first network topology structure using a genetic algorithm combined with the sorting result includes: Step A1: Randomly generate N binary matrices according to the encoding method to form an initial population, where each binary matrix represents an industrial network topology; Step A2: Calculate the fitness value of each individual in the population according to the designed fitness function, where each individual represents a possible topological structure; Step A3: Use roulette wheel selection or tournament selection to select individuals with higher fitness from the current population to enter the next generation population; Step A4: Perform crossover operation on the selected individuals according to the set crossover probability; Step A5: Mutate the individual's genes based on the mutation probability; Step A6: Repeat steps A2-A5 until the set number of iterations is reached or the stopping condition is met, at which point the topology corresponding to the individual with the highest fitness in the population is the first network topology.

3. The method according to claim 2, wherein: The encoding method of the genetic algorithm includes: using binary encoding to represent the industrial network topology. Assume that there are n nodes in the industrial network, then an n×n binary matrix T can be constructed, and the elements T in the matrix are ij , when T ij =1, indicating that there is a connection between node i and node j; when T ij =0, it means there is no connection between node i and node j.

4. The method according to claim 3, wherein: A particle swarm optimization algorithm is used in combination with the sorting result to generate a second network topology structure, including: Step B1: Randomly generate M n×n real number matrices as the initial position of the particle, and initialize the particle's velocity matrix V. The velocity matrix elements V ij The value range of is determined according to the actual situation, generally [-1,1]; Step B2: Calculate the fitness value of each particle according to the designed fitness function; Step B3: Compare the fitness value of each particle's current position with the fitness value of its own historical optimal position. If the current position is better, update the individual optimal position; compare the fitness values ​​of all particles, find the particle with the highest fitness, and set its position as the global optimal position; Step B4: Update the particle's velocity and position according to the following formula: ; Among them, t represents the current iteration number, r1 ij and r2 ij is a random number in the interval [0,1], P ij represents the individual historical optimal position of particle i, G ij Represents the global optimal position, constrains the updated position, and ensures that the matrix elements are in the range [0,1]; Step B5: Repeat steps B2-B4 until the maximum number of iterations is reached or the stopping condition is met, at which point the topology structure corresponding to the global optimal position is the first network topology structure.

5. The method according to claim 4, characterized in that: The particle representation of the particle swarm optimization algorithm includes: representing each particle as an n×n real number matrix X, where n is the number of industrial network nodes, and the matrix element X ij Indicates the possibility or weight of the connection between node i and node j, with a value range of [0,1]. ij When it is close to 1, it means that the possibility of connection between node i and node j is high; when X ij When it is close to 0, it means that the connection possibility is small.

6. The method according to claim 5, characterized in that: The fitness function design of the genetic algorithm or particle swarm optimization algorithm includes: assuming that R represents the network reliability index, T represents the real-time index, S represents the scalability index, and C represents the cost-effectiveness index, and the value range of each index is normalized to be within the interval [0,1]. The fitness function formula is: Among them, w1, w2, w3, and w4 are weight coefficients, and w1+w2+w3+w4=1. The setting of weight coefficients needs to be adjusted according to the actual needs of the industrial network.

7. The method according to claim 6, characterized in that: The dynamic invocation of the topology reconstruction strategy based on the business type and network status changes includes: network status-based conversion: real-time monitoring of industrial network status parameters. When the bandwidth utilization of a certain link in the network continuously exceeds a threshold, and the delay and packet loss rate increase significantly, it is determined that the current topology cannot meet the network requirements. According to a preset conversion strategy, the first network topology is converted to a corresponding topology; During the conversion process, first back up the current network configuration and data, then gradually adjust the connection relationship of network devices, and perform network testing at the same time to ensure the continuity and stability of network services during the conversion process.

8. The method according to claim 7, wherein: The performing topology conversion on the first network topology to obtain a second network topology including a ring topology and a star topology, and realizing adaptive optimization of the topology, includes: The ring topology is generated as follows: in the binary encoding of the genetic algorithm, a continuous sequence of nodes is mandatory, so that these nodes are connected in sequence to form a ring, and each node is connected only to adjacent nodes on the ring, wherein the connection element corresponding to the central node and other nodes in the matrix is ​​1, and the other elements are 0. In the particle swarm optimization algorithm, the initial position and update rules of the particles are set so that the connections between the nodes tend to form a ring structure, and when updating the particle position, the probability of increasing the connection weight of the adjacent nodes is increased; The star topology is generated by specifying a central node in the binary matrix encoding of the genetic algorithm so that all other nodes are connected only to the central node, wherein the connection elements corresponding to the central node and other nodes in the matrix are 1, and the other elements are 0. In the particle swarm optimization algorithm, when initializing particles, the connection weights of most nodes are concentrated on a specific node to form the prototype of the star structure, and the stability of this connection pattern is maintained in subsequent iterations. The second network topology also includes a tree topology and a mesh topology; The tree topology is: defining the connection relationship between the parent node and the child node through hierarchical coding; The mesh topology is as follows: connection constraints are relaxed, connections are allowed to be established between more nodes, and the indicators in the fitness function are used to guide the formation of a mesh structure that needs to be converted.

9. A business topology automation construction system based on an industrial environment, characterized by: It includes data acquisition module, node configuration module, industrial network topology analysis module, industrial business topology map construction module and automatic response module; The data acquisition module is used to collect sensor data, control system data and network communication data of industrial equipment, integrate and process the collected data, and build a real-time database of equipment nodes; The node configuration module is used to identify the logical relationship between devices in the device node real-time database based on a neural network, and dynamically sort the node priorities, wherein each device is a node, each node has a corresponding weight value, and dynamic priority sorting is performed according to the size of the weight value to obtain a sorting result; The industrial network topology analysis module is configured to generate a first network topology using a genetic algorithm or a particle swarm optimization algorithm in combination with the sorting result, dynamically invoke a topology reconstruction strategy based on changes in service type and network status, and perform a topology transformation on the first network topology to obtain a second network topology including a ring topology and a star topology, thereby achieving adaptive optimization of the topology; The industrial service topology map construction module is used to select the first or second network topology structure according to actual network requirements and construct it into an industrial service topology map by combining a graph theory algorithm; The automatic response module is used to monitor the operating status and network connection status of the equipment in real time by combining regular polling and event triggering. When a new device is detected to be connected to the network, the topology update process is automatically triggered to collect the data of the new device and add it to the industrial business topology map to achieve topology update.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

Images