AGV (Automatic Guided Vehicle) integrated noise sensing adaptive clustering method for intelligent factory

Abstract

The invention discloses an AGV (Automatic Guided Vehicle) integrated noise sensing adaptive clustering method for an intelligent factory, which relates to the technical field of wireless communication and comprises the key steps of constructing an energy model to count energy consumption of nodes in a wireless communication process, modeling industrial wireless noise, selecting a dynamic cluster head, dynamically clustering, transmitting hierarchical data and the like. According to the scheme, by constructing the noise interference model conforming to the factory environment characteristics and optimizing the cluster head selection mechanism and the data transmission strategy based on the noise interference model, the stability of the IWSNs is improved, the energy consumption distribution is optimized, the authenticity of wireless communication simulation is enhanced, the optimized routing protocol can more accurately reflect the actual application scene, and the applicability of the routing protocol in the IWSNs is improved.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to an AGV integrated noise perception adaptive clustering method for smart factories. Background Art

[0002] With the development of Industry 4.0 and smart manufacturing, industrial wireless sensor networks (IWSNs) have been widely used in smart factories. IWSNs integrate sensing technology, wireless communication, and intelligent data processing, enabling real-time monitoring of the production environment and equipment status, thereby improving the level of industrial automation. Nodes in IWSNs are divided into sensor nodes and sink nodes. Sensors are primarily responsible for collecting environmental parameters (such as temperature, humidity, vibration, gas concentration, etc.) and equipment operating status information, and transmit this data to sink nodes via wireless communication technology. Sink nodes compress the received multi-source data and upload the processed data to an external data processing center. In smart factories, sensors can be strategically deployed around production workshops, warehouses, and industrial equipment based on specific production needs, forming a real-time monitoring and management network covering the entire production environment.

[0003] Driven by Industry 5.0, the integration of intelligent systems into industrial environments has become increasingly prominent. Automated Guided Vehicles (AGVs), a crucial component of intelligent manufacturing, have evolved from traditional material transport tools into mobile auxiliary nodes integrating data processing and wireless communication capabilities. However, the complexity of industrial environments poses significant challenges to the design of routing protocols for IWSNs.

[0004] Industrial environments are subject to significant noise interference, such as electromagnetic interference and mechanical vibration, which significantly impacts wireless signal stability. However, current simulation studies mostly assume ideal communication environments and fail to fully consider the noise characteristics of industrial environments, resulting in poor performance of optimization algorithms in practical applications. Furthermore, IWSN sensor nodes typically rely on batteries for power, and these nodes are widely distributed and difficult to access in some areas, making battery replacement expensive. Effectively managing node energy is a key task in extending the lifecycle of IWSN networks.

[0005] Traditional data transmission methods often employ clustering structures, with a cluster manager (CD) responsible for aggregating and forwarding data within the cluster to reduce direct communication overhead for cluster member (CM) nodes. However, existing cluster head selection methods often rely on fixed parameter optimization and fail to fully account for the dynamic changes in the industrial environment. This results in excessive energy consumption in some cluster head nodes, impacting the overall stability of the network. Furthermore, existing data transmission strategies fail to fully utilize intelligent devices within the factory, and sensor nodes still bear excessive communication tasks, increasing energy consumption. Therefore, constructing a wireless noise model tailored to the industrial environment and optimizing the cluster head selection mechanism and data transmission strategy based on this model is key to improving the stability and optimizing energy consumption distribution of IWSNs. Summary of the Invention

[0006] The purpose of the present invention is to extend the life cycle of IWSNs in smart factories by introducing an AGV integrated noise-aware adaptive clustering method (A-INAC) for smart factories. This method can enhance the authenticity of simulation experiments, optimize network energy balance and reduce communication energy consumption. The present invention combines energy models to count the energy consumption of nodes during wireless communication. Aiming at the noise interference problem that is prevalent in smart factories, the present invention designs an industrial wireless noise model (IWNM) according to the characteristics of the industrial environment to simulate the impact of noise in the industrial environment on wireless communication, thereby improving the authenticity of simulation experiments. On this basis, A-INAC optimizes the routing protocol from three aspects: CDs selection, dynamic clustering and hierarchical transmission.

[0007] To achieve the above objectives, the present invention provides the following technical solution: an AGV integrated noise perception adaptive clustering method for smart factories, comprising the following steps:

[0008] (a) Construct an energy model to calculate the energy consumption of nodes during wireless communication;

[0009] (b) Industrial wireless noise modeling: Based on the noise interference characteristics in industrial environments, an industrial wireless noise model (IWNM) is constructed to dynamically adjust the received power of wireless signals;

[0010] (c) Dynamic cluster head selection: By calculating the competition factor, the node with high residual energy, close to the base station (BS) and close to the AGV is selected as the cluster head (CD). The competition factor formula is:

[0011] CF(i)=w1×h(E i )+w2×f(d i,BS )+w3×g(d i,AGV );

[0012] Among them, w1w2 and w3 are the weight coefficients of each factor in the formula, and w1+w2+w3=1, where h(Ei ) is the energy factor, f(d i,BS ) is the distance factor from node to BS, g(d i,AGV ) is the distance factor from the node to the AGV;

[0013] (d) Dynamic clustering: Ordinary nodes calculate the metric value based on the received cluster head announcement message (ADV):

[0014]

[0015] Where a and b are the weights of the distance and energy factors, respectively, satisfying the formula a+b=1. i,CD represents the distance from a node to its CD, N represents all nodes deployed in the network, d i,j represents the distance between node i and node j;

[0016] Select the optimal cluster head to join, and the ordinary node selects the cluster head with the smallest metric value to join, and sends a join request (JOIN-REQ) to the cluster head;

[0017] Dynamic weight adjustment: Dynamically adjust the weights of factors in the selection metric formula based on the relationship between the energy status in the network and the balance threshold;

[0018] (e) Hierarchical data transmission: Data is divided into emergency data and regular monitoring data. Emergency data is directly transmitted from the current node to the BS, and regular monitoring data is forwarded through AGV relays. AGVs close to the BS are preferentially selected as relay nodes.

[0019] In order to further optimize the present invention, the following technical solutions may be preferably used:

[0020] Preferably, the energy model in step a adopts a first-order radio model.

[0021] The energy model includes:

[0022] Sending energy consumption:

[0023] Receiving energy consumption; E rx (L) = L × E elec ;

[0024] Where, E elec The energy consumed by the transmitter or receiver when processing a unit of data; e fs With e amp represent the energy consumption coefficients under free space propagation and amplifier models respectively;

[0025] d0 is the distance threshold, where

[0026] Data fusion energy consumption: Ed =L*d d .

[0027] Preferably, the industrial wireless noise model (IWNM) generates noise by the following steps:

[0028] (i) Generate random variables: Generate two independent uniform random variables U1 and U2 from the interval [0,1]:

[0029] U1,U2~U(0,1);

[0030] (ii) Convert the uniformly distributed random variable to a standard normal distribution:

[0031]

[0032] Among them, Z0 and Z1 both obey the standard normal distribution N(0,1).

[0033] (iii) Calculate the standard deviation of noise: The standard deviation of noise σ is defined as:

[0034]

[0035] in is a scaling factor related to the specific factory environment.

[0036] (iv) Generate Gaussian noise and calculate Gaussian noise using the standard normal distribution variable Z:

[0037]

[0038] Where μ is the mean of the noise, which is set to 0 to represent zero-mean Gaussian noise, and Z is Z0 or Z1, which are used alternately to ensure that the generated noise satisfies the Gaussian distribution;

[0039] (v) Adjust the signal power. The signal power Pn after adding IWNM is the combined effect of the initial received power Pr and the noise N, and is expressed as:

[0040] Pn=g×Pr+y×N,

[0041] Where γ and ψ are the noise and signal power coefficients, respectively, reflecting their impact on Pn.

[0042] Preferably, the energy factor h(E i ), BS distance factor f(d i,BS ) and AGV distance factor g(d i,AGV ) are defined as:

[0043]

[0044]

[0045] Among them, E i is the current remaining energy of node i, E init is the initial energy of the node, d i,BS represents the distance from node i to BS, S is the set of all surviving nodes in the current network, d i,j is the Euclidean distance between two nodes, G A is the set of AGVs adjacent to node i, d i,AGV is from node i to set G A The shortest distance.

[0046] Preferably, the weight adjustment mechanism in the dynamic clustering is: when the network average energy exceeds the threshold, the distance between the node and the CD is minimized to reduce the transmission energy consumption; when the network average energy is lower than the threshold, the weight of the remaining energy is increased to protect the low-energy nodes, delay node failure, and optimize the energy distribution of the entire network.

[0047] Preferably, in the hierarchical data transmission of step (e), the transmission path of the periodic monitoring data is:

[0048] (i) The cluster head (CD) sends data to the nearest AGV;

[0049] (ii) The AGV searches for an AGV closer to the BS along the direction of the BS as a relay node to transmit data;

[0050] (iii) If there is no better AGV, it is directly transmitted to the BS.

[0051] Preferably, the calculation formula of the cluster head selection threshold T(i) is:

[0052]

[0053] Where P represents the percentage of CD nodes in all nodes; r is the number of the current execution round; G is the set of nodes that have not been selected as CD in the first 1 / P rounds.

[0054] Preferably, the emergency data includes node failure alarms and communication interruption signals, which have a higher transmission priority than regular monitoring data and are directly transmitted to the BS using a single hop.

[0055] Preferably, the periodic monitoring data comprises aggregated environmental parameters or periodic status reports collected by the CD.

[0056] This paper significantly improves the performance of industrial wireless sensor networks (IWSNs) in smart factories through the innovative design of an industrial wireless noise model (IWNM), a dynamic cluster head selection mechanism, and a hierarchical data transmission strategy. The specific technical advantages are as follows:

[0057] (1) Significantly enhanced simulation realism**

[0058] Industrial Wireless Noise Model (IWNM): Generates Gaussian noise consistent with factory environments through Box-Muller transform, dynamically adjusts received signal power, and simulates actual noise sources such as electromagnetic interference and mechanical vibration, minimizing the error between simulation results and the real environment.

[0059] Dynamic noise calibration: The noise standard deviation is dynamically calculated based on real-time received power to adapt to different factory scenarios (such as the difference in noise intensity between foundries and assembly workshops), improving the practical applicability of the routing protocol.

[0060] (2) Network energy consumption balance optimization

[0061] Dynamic cluster head selection mechanism: This mechanism uses competition factors to comprehensively evaluate node energy, distance from the base station (BS) and AGV to prevent low-energy nodes from being selected as cluster heads (CDs), thereby improving the balance of cluster head energy consumption distribution.

[0062] Dynamic weight adjustment: In the dynamic clustering stage, the weights are automatically switched according to the relationship between the average network energy and the threshold, extending the life of low-energy nodes and the overall life cycle of the network.

[0063] (3) Improved data transmission efficiency and reliability

[0064] Data classification strategy: direct transmission of urgent data to BS: to avoid transmission delay of critical information such as fault alarms;

[0065] Periodic data AGV relay: Leveraging AGV mobility to dynamically optimize relay paths, reduce sensor node communication distances, and reduce energy consumption by 25%;

[0066] AGV-assisted transmission: AGV acts as a high-energy relay node, taking on the task of regular monitoring and data transmission, reducing the load on sensor nodes and avoiding premature energy exhaustion.

[0067] (4) Enhanced network stability and scalability

[0068] Cluster head load balancing: By dynamically adjusting the cluster head selection threshold and metric weight, the cluster head distribution uniformity is improved, avoiding network partitioning caused by cluster head overload in local areas;

[0069] Adaptive topology management: AGVs dynamically adjust their positions based on network status (e.g., moving closer to low-energy cluster head areas), improving relay coverage and enabling network capacity scalability to support large-scale deployments.

[0070] (5) Cost-effectiveness and industrial adaptability

[0071] Zero additional hardware cost: Fully reuse the factory's existing AGV equipment as relay nodes, saving additional deployment costs;

[0072] Wide compatibility: supports mainstream industrial communication protocols (such as ZigBee and LoRa), and is suitable for various scenarios such as discrete manufacturing and process industries;

[0073] Maintenance convenience: Through dynamic energy monitoring and AGV autonomous path planning, the frequency of node maintenance is reduced and operation and maintenance costs are significantly reduced.

[0074] Through systematic innovations in noise modeling, dynamic cluster management, and AGV collaborative transmission, this invention solves the problems of uneven energy consumption, inefficient transmission, and simulation distortion in industrial wireless sensor networks in complex environments, providing a highly reliable and low-cost wireless monitoring solution for smart factories. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] Figure 1 This is a schematic diagram of the overall technical solution flow in Example 1;

[0076] Figure 2 Schematic diagram of the energy consumption model in Example 1;

[0077] Figure 3 This is the A-INAC routing flow chart in Example 1. DETAILED DESCRIPTION

[0078] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0079] Example 1:

[0080] This paper proposes an AGV integrated noise perception adaptive clustering method (A-INAC) for smart factories. The A-INAC technical solution process is as follows: Figure 1 As shown, it mainly includes the following modules:

[0081] (1) Energy model

[0082] IWSNs mainly rely on wireless communication between nodes for signal transmission. The present invention adopts the first-order radio model shown in Figure 2 to describe the energy consumption of sensor nodes in the network.

[0083] The energy consumed by a node when sending and receiving Lbit data is shown in formula (1) and (2):

[0084]

[0085] E rx (L) = L × Eelec (2)

[0086] Where, E elec The energy consumed by the transmitter or receiver when processing a unit of data; e fs With e amp represent the energy consumption coefficients under free space propagation and amplifier models respectively; d0 is the distance threshold, calculated as in formula (3).

[0087]

[0088] Since nodes in IWSNs have data fusion capabilities, their energy consumption characteristics still need further study. Each CD processes the redundant data sent by CMs through data fusion technology. The energy consumed by the CD to fuse Lbit of data can be expressed by equation (4).

[0089] E d =L*d d (4)

[0090] (2) Industrial Wireless Noise Model (IWNM)

[0091] To enhance the realism of simulation experiments, this paper designs a randomized industrial wireless noise model (IWNM). This model dynamically adjusts the received power of wireless signals by simulating the noise characteristics of industrial environments. Specifically, the model dynamically calculates the signal power under noise interference by comprehensively considering factors such as the receive threshold (RxThresh), signal-to-noise ratio (SNR), and inter-node wireless communication power (Pr). The introduction of the IWNM can more accurately reflect the fluctuating characteristics of wireless communications in industrial environments, providing a more realistic simulation environment for subsequent route optimization.

[0092] This model uses Box-Muller transform to generate noise N that conforms to Gaussian distribution. The steps for generating noise are as follows:

[0093] Generate random variables: Generate two independent uniform random variables U1 and U2 from the interval [0,1]:

[0094] U1,U2~U(0,1)(5)

[0095] Calculate the standard normal distribution of random variables: Convert uniform distribution random variables to standard normal distribution:

[0096]

[0097] Among them, Z0 and Z1 both obey the standard normal distribution N(0,1).

[0098] Calculate the standard deviation of the noise: The standard deviation σ of the noise is dynamically determined by formula (8).

[0099]

[0100] in is a scaling factor related to the specific factory environment.

[0101] Calculate the final noise: Calculate Gaussian noise using the standard normal distribution variable Z:

[0102]

[0103] Where μ is the mean of the noise, which is set to 0 to represent zero-mean Gaussian noise. Z takes Z0 and Z1, which are used alternately to ensure that the generated noise satisfies the Gaussian distribution.

[0104] Calculate the signal power after adding noise: Calculate the signal power after adding noise. The signal power Pn after adding IWNM is the combined effect of the initial received power Pr and the noise N, expressed as Equation (10).

[0105] Pn=g×Pr+y×N(10)

[0106] Where γ and ψ are the noise and signal power coefficients, respectively, reflecting their impact on Pn.

[0107] (3) Routing key process

[0108] The routing scheme of this invention consists of three key processes: CD selection, dynamic clustering, and hierarchical transmission. During the CD selection phase, A-INAC calculates a competition factor based on factors such as the node's remaining energy, the node's distance from the base station, and the node's distance from the nearest AGV. Selection as a CD is determined by comparing a random number with the node's threshold. Selected CD nodes broadcast ADV messages. Standard nodes then select the optimal cluster head based on a metric calculated based on signal strength and remaining energy. Cluster head load balancing is achieved through a dynamic weighting mechanism. During the hierarchical transmission phase, direct or multi-hop transmission is selected based on the cluster head's communication status, effectively reducing network energy consumption.

[0109] The flowchart of the A-INAC algorithm is as follows Figure 3 shown.

[0110] ①CDs selection

[0111] In the CDs selection phase, this paper proposes a dynamic selection mechanism based on competitive factors. This mechanism comprehensively considers multiple key attributes of nodes and assigns different competitive strengths to different nodes[8], thereby ensuring the rationality of the selected CDs.

[0112] Calculate the competition factor: Calculate the competition factor based on the remaining energy of the node, the distance between the node and the BS, and the distance between the node and the nearest AGV. The calculation method of the competition factor is as shown in Equation (11).

[0113] CF(i) = w1 × h(E i ) + w2 × f(d i,BS ) + w3 × g(d i,AGV )(11)

[0114] Where w1, w2, and w3 are the weight coefficients of each factor in the formula, and w1 + w2 + w3 = 1. The energy factor h(E i ), the BS distance factor f(d i,BS ), and the AGV distance factor g(d i,AGV ) are defined as Formula (12), Formula (13), and Formula (14) respectively.

[0115]

[0116] Where E i is the current remaining energy of node i, and E init is the initial energy of the node. d i,BS represents the distance from node i to the BS, S is the set of all surviving nodes in the current network, and d i,j is the Euclidean distance between two nodes. G A is the set of AGVs adjacent to node i, and d i,AGV is the shortest distance from node i to the set G A .

[0117] Determine the threshold: The formula for calculating the threshold containing the competition factor after optimization is expressed as Equation (15):

[0118]

[0119] Where P represents the percentage of CDs among all nodes; r is the current execution round; G is the set of nodes not selected as CDs in the previous 1 / P rounds.

[0120] Select the cluster head: Each node generates a random number R(i) and compares it with the threshold T(i). If R(i) < T(i), then the node is selected as the CD and broadcasts a cluster head announcement message (ADV).

[0121] ② Dynamic clustering

[0122] In the dynamic clustering stage, ordinary nodes select the optimal cluster head to join according to the received ADV message. The present invention proposes a clustering mechanism based on metric values to ensure that the formation of clusters can adapt to changes in the network state.

[0123] Calculate the metric value: Each ordinary node calculates a metric value for each ADV message received. The formula for selecting the metric is equation (16).

[0124]

[0125] Where a and b are the weights of the distance and energy factors, respectively, satisfying the formula a+b=1. i,CD represents the distance from a node to its CD, N represents all nodes deployed in the network, d i,j Represents the distance between node i and node j.

[0126] Select the optimal cluster head: Ordinary nodes select the cluster head with the smallest metric value to join and send a join request (JOIN-REQ) to the cluster head.

[0127] Dynamic Weight Adjustment: To further optimize energy distribution during clustering, this invention introduces a balance threshold mechanism. Based on the relationship between the network's energy status and the balance threshold, the weights of the factors in the selection metric are dynamically adjusted. Specifically, when the average network energy exceeds the threshold, the distance between the node and the CD is minimized to reduce transmission energy consumption. Conversely, when the average network energy exceeds the threshold, the weight of the remaining energy is increased to protect low-energy nodes, thereby delaying node failure and optimizing the energy distribution of the entire network.

[0128] ③ Hierarchical transmission

[0129] Modern AGVs are equipped with advanced communication modules and data processing capabilities, enabling seamless data integration into IWSNs. Their mobility enables dynamic relocation based on network topology and transmission requirements, optimizing relay efficiency. This paper introduces a hierarchical transmission strategy during the data transmission phase. This strategy categorizes data into emergency data and regularly monitored data based on their communication status, and employs different transmission strategies to reduce network energy consumption.

[0130] Data classification: The cluster head divides the data into two categories according to the communication status of the data.

[0131] Emergency data: such as alerts generated when a node fails or information carried by unsuccessful cluster entry.

[0132] Regular monitoring data: This includes summarized environmental parameters or regular status reports collected by the CD.

[0133] Hierarchical transmission: Different types of data use different transmission strategies.

[0134] Urgent data transmission: Urgent data is transmitted directly from the current node to the BS to ensure the timeliness of the data.

[0135] Regularly monitor data transmission: The CD first forwards the data to the AGV closest to the node, then gradually searches for an AGV closer to the BS as a relay node to transmit the data. If no better AGV is found, the data is transmitted to the BS.

[0136] The hierarchical transmission method designed in this paper can utilize the existing AGVs in the factory for data transmission without increasing factory costs, maximizing the use of available resources. This strategy effectively balances the energy consumption of CD by offloading part of the transmission workload to AGVs outside the network.

[0137] The core advantages of this solution are:

[0138] Factory noise interference modeling: In response to the complex noise in industrial environments, this paper designs a noise interference model based on the actual factory situation to simulate the impact of equipment on wireless communication, making the simulation closer to reality.

[0139] Data classification strategy: classify data according to network communication status to reduce unnecessary communication overhead and optimize data transmission efficiency.

[0140] Smart device-assisted data transmission: When transmitting regular monitoring data, existing smart devices in the factory (such as AGVs) are used as relays to reduce the communication burden of ordinary sensor nodes and extend the network lifecycle.

[0141] The term "comprise" or any other similar term is intended to cover non-exclusive inclusion, such that a process, article, or apparatus / device that comprises a list of elements includes not only those elements but also other elements not expressly listed or inherent to such process, article, or apparatus / device.

[0142] Thus far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art may make equivalent changes or substitutions to the relevant technical features, and the technical solutions after such changes or substitutions will fall within the scope of protection of the present invention.

Claims

1. An AGV integrated noise perception adaptive clustering method for smart factories, characterized by: The following steps are involved: (a) Construct an energy model to calculate the energy consumption of nodes during wireless communication; (b) Industrial wireless noise modeling: Based on the noise interference characteristics in industrial environments, an industrial wireless noise model (IWNM) is constructed to dynamically adjust the received power of wireless signals; (c) Dynamic cluster head selection: By calculating the competition factor, the node with high residual energy, close to the base station (BS) and close to the AGV is selected as the cluster head (CD). The competitive factor formula is defined as: CF(i) = w1×h(E i )+w2×f(d i,BS )+w3×g(d i,AGV ); Among them, w1 w2 and w3 are the weight coefficients of each factor in the formula, and w1+w2+w3=1, where h(E i ) is the energy factor, f(d i,BS ) is the distance factor from node to BS, g(d i,AGV ) is the distance factor from the node to the AGV; (d) Dynamic clustering: ordinary nodes receive cluster head announcement messages (ADV), The calculated measure is defined as: Where a and b are the weights of the distance and energy factors, respectively, satisfying the formula a+b=1. i,CD represents the distance from a node to its CD, N represents all nodes deployed in the network, d i,j represents the distance between node i and node j; Select the optimal cluster head to join, and the ordinary node selects the cluster head with the smallest metric value to join, and sends a join request (JOIN-REQ) to the cluster head; Dynamic weight adjustment: Dynamically adjust the weights of factors in the selection metric formula based on the relationship between the energy status in the network and the balance threshold; (e) Hierarchical data transmission: Data is divided into emergency data and regular monitoring data. Emergency data is directly transmitted from the current node to the BS, and regular monitoring data is forwarded through AGV relays. AGVs close to the BS are preferentially selected as relay nodes.

2. The AGV integrated noise perception adaptive clustering method for smart factories according to claim 1 is characterized in that: The energy model in step a adopts a first-order radio model. The energy model includes: Sending energy consumption: Receiving energy consumption; E rx (L) = L × E elec ; Where, E elec The energy consumed by the transmitter or receiver when processing a unit of data; e fs With e amp represent the energy consumption coefficients under free space propagation and amplifier models respectively; d0 is the distance threshold, where Data fusion energy consumption: E d =L*d d .

3. The AGV integrated noise perception adaptive clustering method for smart factories according to claim 1 is characterized in that: The Industrial Wireless Noise Model (IWNM) generates noise through the following steps: (i) Generate random variables: Generate two independent uniform random variables U1 and U2 from the interval [0,1]: U1,U2~U(0,1); (ii) Convert the uniformly distributed random variable to a standard normal distribution: Among them, Z0 and Z1 both obey the standard normal distribution N(0,1); (iii) Calculate the standard deviation of noise: The standard deviation of noise σ is defined as: in is a scaling factor related to the specific factory environment; (iv) Generate Gaussian noise and calculate Gaussian noise using the standard normal distribution variable Z: Where μ is the mean of the noise, which is set to 0 to represent zero-mean Gaussian noise, and Z is Z0 or Z1, which are used alternately to ensure that the generated noise satisfies the Gaussian distribution; (v) Adjust the signal power. The signal power Pn after adding IWNM is the combined effect of the initial received power Pr and the noise N, and is expressed as: Pn=g×Pr+y×N, Where γ and ψ are the noise and signal power coefficients, respectively, reflecting their impact on Pn.

4. The AGV integrated noise perception adaptive clustering method for smart factories according to claim 1 is characterized in that: The energy factor h(E i ), BS distance factor f(d i,BS ) and AGV distance factor g(d i,AGV ) are defined as: Among them, E i is the current remaining energy of node i, E init is the initial energy of the node, d i,BS represents the distance from node i to BS, S is the set of all surviving nodes in the current network, d i,j is the Euclidean distance between two nodes, G A is the set of AGVs adjacent to node i, d i,AGV is from node i to set G A The shortest distance.

5. The AGV integrated noise perception adaptive clustering method for smart factories according to claim 1 is characterized in that: The weight adjustment mechanism in the dynamic clustering is as follows: when the average network energy exceeds a threshold, the distance between the node and the CD is minimized to reduce transmission energy consumption; when the average network energy is lower than the threshold, the weight of the remaining energy is increased to protect low-energy nodes, delay node failure, and optimize the energy distribution of the entire network.

6. The AGV integrated noise perception adaptive clustering method for smart factories according to claim 1 is characterized in that: In the step (e) of hierarchical data transmission, the transmission path of the periodic monitoring data is: (i) The cluster head (CD) sends data to the nearest AGV; (ii) The AGV searches for an AGV closer to the BS along the direction of the BS as a relay node to transmit data; (iii) If there is no better AGV, it is directly transmitted to the BS.

7. The AGV integrated noise perception adaptive clustering method for smart factories according to claim 1 is characterized in that: The calculation formula of the cluster head selection threshold T(i) is defined as: Where P represents the percentage of CD nodes in all nodes; r is the number of the current execution round; G is the set of nodes that have not been selected as CD in the first 1 / P rounds.

8. The AGV integrated noise perception adaptive clustering method for smart factories according to claim 1 is characterized in that: The emergency data includes node failure alarms and communication interruption signals, and has a higher transmission priority than regular monitoring data, and is directly transmitted to the BS using a single hop.

9. The AGV integrated noise perception adaptive clustering method for smart factories according to claim 1 is characterized in that: The periodic monitoring data may include summarized environmental parameters or periodic status reports collected by the CD.

Images