A Competitive Energy-Efficient Clustering Method for Wireless Sensor Networks Based on Fuzzy Width Learning

Abstract

This application relates to the field of wireless sensor network clustering optimization, and more specifically, to a wireless sensor network energy efficiency competition clustering method based on fuzzy width learning. The method includes: collecting historical node information of the wireless sensor network; preprocessing the historical node information to generate a historical node dataset; training and testing a fuzzy width learning model based on the historical node dataset to obtain a trained model; generating nodes of the wireless sensor network and their corresponding node information; preprocessing the node information and inputting it into the trained model to obtain the competitiveness value of each node; obtaining the optimal number of cluster heads; selecting cluster head nodes based on the node competitiveness values, and designating other nodes as ordinary nodes; and constructing several clusters in the wireless sensor network for data transmission based on the cluster head nodes and ordinary nodes. This method is used to dynamically select suitable cluster head nodes and optimize the clustering process to meet the real-time and adaptive requirements of wireless sensor networks.

Description

Technical Field

[0001] This invention relates to the field of wireless sensor network clustering optimization, and more specifically, to a wireless sensor network energy efficiency competitive clustering method based on fuzzy width learning. Background Technology

[0002] With the rapid development of IoT technology, Wireless Sensor Networks (WSNs), as a core technology within it, have been widely applied in various fields such as environmental monitoring, military reconnaissance, medical care, and smart homes. A WSN consists of a large number of sensor nodes with sensing, computing, and communication capabilities. These sensor nodes are distributed throughout the monitoring area and collaborate to complete data acquisition, processing, and transmission tasks. The collected data is reported to a base station (BS), enabling data communication between the end user and the WSN. However, since sensor nodes are typically powered by batteries, energy resources are limited and difficult to replenish. Therefore, how to effectively utilize sensor node energy to extend network lifespan has become one of the core issues in WSN research.

[0003] To address the issue of limited energy and avoid the enormous energy consumption caused by direct communication between sensor nodes and the base station (BS), cluster routing is a widely researched and applied method, a cluster-based transmission protocol. Cluster routing divides the wireless sensor network into multiple clusters, each consisting of a cluster head node and several ordinary nodes. The cluster head node (CH) is responsible for aggregating the data collected by the ordinary nodes within the cluster and forwarding it to the base station (BS). This method effectively reduces communication overhead between nodes, thereby lowering energy consumption. To achieve the above effects, the rational selection of the cluster head node for each cluster is a necessary area of ​​research. In existing technologies, a cluster head node (CH) selection protocol based on LEACH low-power adaptive clustering hierarchical routing is disclosed, but it suffers from drawbacks such as randomness and uncertainty in CH selection, and low-energy nodes may be selected as CH, leading to premature energy depletion. Another method is based on a fuzzy logic system for dynamic CH selection, but it relies too heavily on expert experience and performs poorly in dynamic environments. A heuristic method based on chaotic ant colonies is also disclosed for the clustering process, but it easily encounters performance bottlenecks in large-scale problems. Furthermore, other complex deep learning methods cannot meet the real-time requirements of modern wireless sensor networks due to excessive time overhead. Therefore, achieving reasonable cluster head selection and cluster construction in complex dynamic environments remains a challenge in wireless sensor networks. Summary of the Invention

[0004] The present invention aims to overcome at least one of the defects (deficiencies) of the prior art and provide a wireless sensor network energy efficiency competitive clustering method based on fuzzy width learning, which is used to dynamically select suitable cluster head nodes and optimize the clustering process to meet the real-time and adaptive requirements of wireless sensor networks.

[0005] According to a first aspect of this application, a method for energy-efficiency competitive clustering of wireless sensor networks based on fuzzy width learning is provided, the method comprising:

[0006] Collect information on several historical nodes of a wireless sensor network, and preprocess the information on several historical nodes to generate a historical node dataset.

[0007] The fuzzy width learning model is trained and tested based on the historical node dataset to obtain a trained fuzzy width learning model.

[0008] Generate several nodes of the wireless sensor network and their corresponding node information. After preprocessing the node information, input it into the trained fuzzy width learning model. Calculate the competitiveness value of the node corresponding to each node information through the fuzzy width learning model.

[0009] Obtain the attributes of the wireless sensor network, and obtain the optimal number of cluster heads for the wireless sensor network based on the attributes of the wireless sensor network.

[0010] Based on the competitiveness value of each node, a number of cluster head nodes are selected, wherein the number of selected cluster head nodes is equal to the number of optimal cluster heads; nodes not selected as cluster head nodes are set as ordinary nodes.

[0011] Based on the cluster head nodes and the ordinary nodes, several clusters are constructed in the wireless sensor network for data transmission.

[0012] Understandably, introducing a fuzzy width learning model to calculate the competitiveness value of each node in the wireless sensor network, and selecting suitable cluster head nodes by comparing the competitiveness values ​​of each node, and obtaining the optimal number of cluster heads based on the attributes of the wireless sensor network, can make the transmission operation of the entire wireless sensor network more efficient and orderly, while optimizing the clustering process of the wireless sensor network and meeting the real-time and adaptive requirements of the wireless sensor network.

[0013] Optionally, the step of preprocessing the historical node information to generate a historical node dataset includes:

[0014] Data normalization processing is performed on several historical node information to obtain several historical node information with the same dimension.

[0015] Missing value processing is performed on several historical node information with the same dimension to obtain several complete historical node information; wherein the missing value processing is to obtain node information with missing competitiveness value in the several node information with the same dimension, input the node information with missing competitiveness value into a fuzzy logic system for processing to obtain the corresponding competitiveness value, and complete the information of the node with missing competitiveness value according to the obtained competitiveness value.

[0016] The historical node dataset is obtained by converting the data format of several complete historical node information.

[0017] Understandably, performing data normalization, missing value handling, and data format conversion on historical node information can provide a complete and normalized dataset for subsequent training of the fuzzy width learning model, thereby improving the efficiency of the fuzzy width learning model in data processing and obtaining as much training information as possible.

[0018] Optionally, the competitiveness value depends on an influencing factor, which includes one or more of the following: the node's relative remaining energy Energy_Re(i), the node's relative distance to the base station Dist_BS(i), the node's neighbor node density Nei_Dens(i), and the node's historical cluster head state CH_State(i).

[0019] Understandably, it is necessary to train and test the fuzzy width learning model so that the trained fuzzy width learning model can obtain the competitiveness value of a node based on the input node information. This application provides four influencing factors of competitiveness value. When training and testing the fuzzy width learning model, these influencing factors can be used as the influence data of the method for learning to calculate competitiveness value, so that the method for calculating competitiveness value learned by the fuzzy width learning model is more accurate and the obtained competitiveness value is reliable.

[0020] Optionally, the relative remaining energy Energy_Re(i) of the node is obtained by the following formula:

[0021]

[0022] Energy res (i) represents the remaining energy of node i. init (i) represents the initial energy of node i;

[0023] And / or,

[0024] The relative distance Dist_BS(i) from the node to the base station BS is obtained by the following formula:

[0025]

[0026] Where d1, d2, d3, and d4 represent the distances from the four boundary corners of the wireless sensor network scenario to the base station BS, and max(.) is the maximum value function, representing the selection of the distance from d1, d2, d3, and d4 to the base station BS; Dist BS (S i The distance () represents the Euclidean distance from node i to the base station BS, which is obtained through the following calculation formula:

[0027]

[0028] Where x i y i Let x and y represent the x and y coordinates of node i, respectively. BS y BS These represent the horizontal and vertical coordinates of the base station BS, respectively.

[0029] And / or,

[0030] The density of neighbor nodes Nei_Dens(i) of node i is obtained by the following formula:

[0031]

[0032] Where, N w The total number of nodes in the wireless sensor network is represented by ; Neighbor(i) represents the number of neighboring nodes of node i, which is obtained by the following formula:

[0033]

[0034] Where j represents other nodes j in the wireless sensor network besides node i, and d ij R represents the distance between node i and the other nodes j. t This represents the maximum radius distance at which node i transmits relevant information, and δ(.) is an indicator function, where d ij When ≤R, δ=1; otherwise, δ=0;

[0035] And / or,

[0036] The historical cluster head state CH_State(i) of the node is obtained by the following formula:

[0037]

[0038] Where r represents the number of times node i is consecutively selected as the cluster head node.

[0039] Understandably, this application introduces the relative remaining energy of a node, the relative distance between a node and a base station (BS), the density of neighboring nodes of a node, and the methods for obtaining the relative remaining energy of a node, and provides formulas and methodologies for obtaining influencing factors; among them, the historical cluster head state of a node is used as one of the influencing factors in calculating the competitiveness value, which can fully consider the real-time energy state of each node, so that the obtained competitiveness value can better reflect the most realistic state of a node, and make the cluster head nodes selected later more accurate.

[0040] Optionally, the fuzzy width learning model includes an input layer, a fuzzy feature layer, a feature enhancement layer, and an output layer; the fuzzy feature layer includes several fuzzy subsystems.

[0041] The step of training and testing the fuzzy width learning system network model based on the historical node dataset includes:

[0042] The historical node dataset is divided into a training set and a test set;

[0043] Define the number of fuzzy subsystems Q and the number of fuzzy rules K;

[0044] Initialize the coefficient weights of the fuzzy width learning model;

[0045] The training set is input into the fuzzy width learning model through the input layer;

[0046] In the fuzzy feature layer, an intermediate matrix Z is generated through several fuzzy subsystems of the fuzzy width learning model and the training set. n and feature node matrix F n , where n is the total number of training samples in the training set;

[0047] The intermediate matrix Z n Each node feature value is input into the feature enhancement layer, and several enhanced nodes H are obtained through enhanced feature processing. m According to the feature node matrix F n and the aforementioned several enhancement nodes H m Obtain the target value Y, and quickly calculate the weight matrix W based on the target value Y and the pseudo-inverse.

[0048] In the output layer, the preliminary parameters of the fuzzy width learning model are obtained based on the weight matrix W;

[0049] The test set is input into the fuzzy width learning model, and the error is calculated, including the error rate ERR and the mean absolute percentage error MAPE.

[0050] After optimizing the number of fuzzy subsystems Q and the number of fuzzy rules K based on the error rate ERR and mean absolute percentage error MAPE, the fuzzy width learning model is iteratively trained. The initial parameters are optimized to obtain the optimal model parameters, and the training of the fuzzy width learning model is completed.

[0051] Understandably, training and testing the fuzzy width learning model, and iteratively optimizing the parameters of the fuzzy width learning model by calculating relevant errors until a certain threshold is reached, can provide a clear methodology for training the fuzzy width learning model, making the entire training process clearer and more organized.

[0052] Optionally, each training sample in the training set is N×M dimensional data, where N represents the number of training sample groups and M represents the feature information dimension of the training sample, specifically as follows:

[0053] X = (x1, x2, ..., x s ,...,x N ) T ∈R N×M

[0054] Where x s =(x s1 ,x s2 ,…,x sm ,…,x sM ), s=1,2,…,N;

[0055] And / or, the setting of the number of fuzzy subsystems Q and the number of fuzzy rules K specifically includes:

[0056] In the q-th fuzzy subsystem, K is set q The fuzzy rules are represented in the following form:

[0057] If x s1 is and x s2 is … and x sM is

[0058] but

[0059] Where k = 1, 2, ..., K q x sm This represents one of the training samples in the training set. This represents a preset fuzzy set in the fuzzy subsystem. u is an adjustable parameter of the model rule.qg This represents the training sample g in the training set for the fuzzy rule K. q membership degree The intermediate matrix Z n The output of the q-th fuzzy subsystem, where u qg It can be represented as:

[0060]

[0061] Among them, c q c k This represents the cluster center corresponding to the fuzzy rule;

[0062] Then the training sample x s The output Z of the q-th fuzzy subsystem sq Represented as:

[0063]

[0064] in, The output Z for the q-th fuzzy subsystem sq single middle value The weights;

[0065] The intermediate matrix Z of the training set output by the combination of the Q fuzzy subsystems n Represented as:

[0066] Z n =(Z1,Z2,…,Z n )

[0067] The deblurred output F of the blurry feature layer sq Represented as:

[0068]

[0069] in, These are adjustable parameters;

[0070] The defuzzification output F of the q-th fuzzy subsystem q for:

[0071]

[0072] in, diag{.} is a diagonal matrix;

[0073] The output F of the fuzzy feature layer n Represented as:

[0074]

[0075] And / or, the intermediate matrix Z n Each node feature value is input into the feature enhancement layer, and several enhanced nodes H are obtained through enhanced feature processing. m ,include:

[0076] The intermediate matrix Z n Z value of each feature node in I The value H mapped to the augment node J :

[0077]

[0078] Where, ξ J (.) is a mapping function. For random weights, For bias terms;

[0079] Then the several enhancement nodes H of the m features m The output is represented as:

[0080] H m =(H1,H2,…,H m )

[0081] And / or, according to the feature node matrix F n and the aforementioned several enhancement nodes H m To obtain the target value Y, the following steps are required:

[0082] The target value Y output by the output layer is:

[0083]

[0084] Among them, W f W represents the weight coefficients from the blurred feature layer to the output layer. h These are the weight coefficients from the feature enhancement layer to the output layer;

[0085] And / or, the weight matrix W is obtained by fast calculation based on the target value Y and the pseudo-inverse, specifically according to the following formula:

[0086] W=(Bω,H m ) + Y

[0087] Among them, (Bω,H m ) + =((Bω,H m ) T (Bω,H m )) -1 (Bω,H m ) T ;

[0088] And / or, the formula for calculating the Mean Absolute Percentage Error (MAPE) is expressed as follows:

[0089]

[0090] Among them, y v Let the v-th intermediate target true value be the target value Y. For y v The predicted value;

[0091] And / or, the formula for calculating the error rate ERR is expressed as:

[0092]

[0093] Among them, E baseline E represents the baseline error value of the fuzzy width learning model. m0del This represents the error value in the current training process of the fuzzy width learning model.

[0094] Understandably, the above illustrates the calculation formulas for relevant steps during the training or testing process. This provides an accurate calculation method for the entire training process, making the calculations and processing in the training and testing of the fuzzy width model more organized and clear.

[0095] Optionally, the attributes of the wireless sensor network include the total number of nodes in the wireless sensor network, the size of the wireless sensor network, and the average distance from each node in the wireless sensor network to the base station (BS).

[0096] The step of obtaining the optimal number of cluster heads for the wireless sensor network based on its attributes includes:

[0097] Optimal number of cluster heads k opt Specifically, it can be obtained through the following formula:

[0098]

[0099] Where, N w Let d be the total number of nodes in the wireless sensor network, P be the size of the wireless sensor network, and d be the total number of nodes in the wireless sensor network. toBS The average distance from each node in the wireless sensor network to the base station (BS);

[0100] The average distance from each node in the wireless sensor network to the base station (BS) is specifically obtained using the following formula:

[0101]

[0102] Understandably, the required number of cluster heads varies depending on the type of wireless sensor network, including the total number of nodes, range size, and attribute information of each node. Too few cluster heads cannot effectively complete the transmission problem of the network, while too many cluster heads may waste relevant resources. Therefore, it is necessary to obtain the optimal number of cluster heads based on the attributes of each wireless sensor network to make the transmission work of the wireless sensor network more rationally allocated and improve the transmission efficiency.

[0103] Optionally, the step of selecting several cluster head nodes based on the competitiveness value of each node includes:

[0104] Sort all the nodes by competitiveness value from high to low, select the node with high competitiveness as the cluster head node, and the rest of the nodes as ordinary nodes.

[0105] And / or,

[0106] The step of constructing several clusters in the wireless sensor network based on several cluster head nodes and several ordinary nodes for data transmission includes:

[0107] The clusters are constructed using a second cluster head strategy, a sub-cluster head strategy, and a minimum energy consumption criterion strategy.

[0108] The second cluster head strategy involves selecting the node with the highest competitiveness value (excluding the cluster head node) from the constructed cluster group to serve as the second cluster head, and the second cluster head assists the cluster head node in transmission work.

[0109] The secondary cluster head strategy involves selecting a node in the cluster that can replace the cluster head node as a secondary cluster head node. When the cluster head node fails or cannot work properly, the secondary cluster head node takes over as the cluster head node and completes the work of the cluster head node.

[0110] The minimum energy consumption criterion strategy calculates the indirect energy consumption E of data transmission from the ordinary node to the base station BS via the cluster head node when the ordinary node transmits data in the wireless sensor network. idt And the direct energy consumption E of computing data being transmitted directly from the ordinary node to the base station BS. dt And compare the indirect energy consumption E using a cost function. idt and the direct energy consumption E dt The size of the path determines the path with the lowest energy consumption for transmission.

[0111] Understandably, selecting cluster head nodes based on their competitiveness values ​​accurately reflects the node's status and transmission capabilities, thus choosing a more suitable cluster head to ensure the smooth completion of complex transmission tasks undertaken by the cluster head, thereby enabling the entire wireless sensor network to function smoothly. Building clusters in a wireless sensor network requires adhering to a second cluster head strategy, a sub-cluster head strategy, and a minimum energy consumption criterion strategy. The second cluster head strategy ensures the cluster head node's work is completed smoothly and efficiently, reducing its workload. The sub-cluster head strategy ensures rapid response when a cluster head node fails, minimizing the impact of the failure on the overall transmission operation. The minimum energy consumption criterion strategy ensures reduced energy consumption for the entire transmission operation, making it more economical and reliable.

[0112] Optionally, the indirect energy consumption and / or direct energy consumption are affected by the following parameters:

[0113] The distance 'a' from a regular node to the cluster head node is specifically related to: E idt ∝a;

[0114] The distance b from the cluster head node to the base station BS is specifically related to: E idt ∝b;

[0115] The distance c from a regular node to a base station (BS) is specifically related to: E dt ∝c;

[0116] The indirect energy consumption E idt Specifically, it can be obtained through the following formula:

[0117] E idt =l·[(1+μ)E elec +∈ fs (a 2 +μb 2 )]

[0118] Where μ represents the preset data aggregation rate;

[0119] Among them, E elec The energy consumption coefficient of the path, ∈ fs This represents the power amplification factor of the existing free-space model;

[0120] The direct energy consumption E dt Specifically, it can be obtained through the following formula:

[0121] E dt =l·(E elec +∈ fs c 2 )

[0122] The indirect energy consumption E idt and the direct energy consumption Edt The energy consumption difference ΔE is obtained through the following formula:

[0123] △E=E idt -E dt =l·[μE elec +ε fs (a 2 +μb 2 -c 2 )]

[0124] When △E>0, it means E dt Numerically smaller, the ordinary node selects data to be transmitted directly from the ordinary node to the base station (BS) as the transmission path for data transmission; when ΔE < 0, it indicates that E idt Numerically smaller, the ordinary node data is indirectly transmitted from the ordinary node to the base station (BS) via the cluster head node as the transmission path for data transmission.

[0125] Understandably, selecting an appropriate path for information transmission can save energy consumption during transmission, thereby reducing transmission consumption in the entire wireless sensor network. Optionally, the method further includes:

[0126] After the cluster is constructed, the base station (BS) sends an acknowledgment message (BS_info message) to each node in the wireless sensor network, including the ID of the cluster head node, the coordinates of the cluster head node, and the cluster to which the cluster head node belongs; each node starts working after receiving the acknowledgment message and enters the data collection and transmission phase.

[0127] During the data collection and transmission phase, the cluster head node performs data aggregation on the collected data; wherein, the data aggregation adopts the IA data aggregation model, and the cluster head node aggregates and forwards the received data packets according to the data aggregation rate, with the specific calculation formula as follows:

[0128] L agg =L r +μ×L r ×N p

[0129] Among them, L agg L represents the length of the aggregated data packet. r The length of the received data packet is represented by μ, and the data aggregation rate is represented by N. p This indicates the number of data packets received.

[0130] Understandably, this provides a reliable method for data transmission to cluster head nodes, reduces data redundancy, makes the information accessed in the entire transmission process more flexible and concise, and improves the usability of wireless sensor network memory.

[0131] Based on any of the above aspects, this application provides a method for energy efficiency competition clustering of wireless sensor networks based on fuzzy width learning. This method involves collecting historical node information from a wireless sensor network, preprocessing the historical node information to generate a historical node dataset, training and testing a fuzzy width learning model based on the historical node dataset, generating a trained fuzzy width learning model, generating several nodes and corresponding node information for the wireless sensor network, preprocessing the node information, and inputting it into the trained fuzzy width learning model to calculate the competitiveness value of each node. The method also involves acquiring the attributes of the wireless sensor network and determining the optimal number of cluster heads based on these attributes; selecting several cluster head nodes based on the competitiveness value of each node, wherein the number of selected cluster head nodes is equal to the optimal number of cluster heads; nodes not selected as cluster head nodes are designated as ordinary nodes; and constructing several clusters in the wireless sensor network for data transmission based on the cluster head nodes and the ordinary nodes. Compared with the prior art, this invention has the following advantages and technical effects:

[0132] 1) This invention provides a competitive clustering method based on a fuzzy width learning model that can accurately and quickly process the clustering process in the complex and ever-changing environment of wireless sensor networks, thus solving the problem of unreasonable cluster head decision-making.

[0133] 2) The competitiveness calculation of each node based on the fuzzy width learning model proposed in this invention combines the feature information processing capability of the fuzzy system with the fast training capability of the width learning system.

[0134] 3) This invention adaptively determines the optimal cluster head node under the current situation based on the dynamic changes of the network environment, which effectively improves the energy efficiency of the network; in addition, the addition of the "second cluster head" and "sub-cluster head" strategies realizes the balance of network energy efficiency and also provides a fault tolerance mechanism.

[0135] 4) The energy efficiency competitive clustering method for wireless sensor networks proposed in this invention, based on fuzzy width learning, has shown good advantages in energy efficiency, network lifetime and network stability through simulation experiments in multiple environments. Attached Figure Description

[0136] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0137] Figure 1 This embodiment provides a flowchart of a wireless sensor network energy efficiency competitive clustering method based on fuzzy width learning.

[0138] Figure 2 This embodiment provides a flowchart of a method for generating a historical node dataset.

[0139] Figure 3 This embodiment provides a flowchart of a method for training and testing a fuzzy width learning model.

[0140] Figure 4 The diagram shows the average remaining energy of nodes in different embodiments of the present invention and the comparative method provided in this embodiment.

[0141] Figure 5 The graph shows the fluctuation of the average remaining energy performance of the nodes in different embodiments of the methods provided in this embodiment. Detailed Implementation

[0142] The accompanying drawings are for illustrative purposes only and should not be construed as limiting the scope of this application. To better illustrate the following embodiments, some components in the drawings may be omitted, enlarged, or reduced, and do not represent the actual dimensions of the product; it is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.

[0143] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0144] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0145] With the rapid development of IoT technology, wireless sensor networks, as its core technology, have been widely applied in various fields. A wireless sensor network consists of a large number of sensor nodes with sensing, computing, and communication capabilities. These nodes are distributed across the monitoring area, collaboratively completing data acquisition, processing, and transmission tasks, with the data ultimately reported to the base station (BS). However, sensor nodes have limited and difficult-to-replenish energy; therefore, how to effectively utilize energy to extend network lifetime has become a core research issue.

[0146] To address the issue of limited energy, clustering routing methods are widely used. This method divides the wireless sensor network into multiple clusters, each consisting of a cluster head node (CH) and several ordinary nodes. The CH is responsible for aggregating and forwarding data within its cluster to the base station (BS), thereby reducing communication overhead and energy consumption. The appropriate selection of cluster head nodes is crucial, but existing technologies have shortcomings: methods based on LEACH-based low-power adaptive clustering hierarchical routing protocols suffer from randomness and uncertainty; fuzzy logic systems rely on expert experience and have poor dynamic performance; chaotic ant colony methods have performance bottlenecks in large-scale problems; and deep learning methods are time-consuming and difficult to meet real-time requirements. Therefore, how to achieve reasonable cluster head selection and cluster construction in complex dynamic environments remains a major challenge for wireless sensor networks.

[0147] This embodiment provides a technical solution that can solve the above problems. The specific implementation of this application will be described in detail below with reference to the accompanying drawings.

[0148] like Figure 1 As shown, this embodiment provides a fuzzy width learning-based energy efficiency competitive clustering method for wireless sensor networks, which may include the following steps:

[0149] S110. Collect several historical node information of the wireless sensor network, and preprocess the several historical node information to generate a historical node dataset.

[0150] In this embodiment, the historical node information includes several historical nodes in the historical transmission work of the wireless sensor network. The historical nodes also include corresponding historical node information, which includes the node's unique ID, the node's location information in the wireless sensor network, the node's energy information, and the node's status information. Understandably, each node has a corresponding node device to complete the corresponding transmission work, and the historical node information is stored in the node device for retrieval or use.

[0151] Specifically, such as Figure 2 As shown, the step of preprocessing the historical node information to generate a historical node dataset includes the following steps:

[0152] S111. Perform data normalization processing on several historical node information to obtain several historical node information with the same dimension.

[0153] In this embodiment, the data normalization process refers to normalizing all historical node information so that it is distributed in (0,1), so that historical node information of different dimensions has the same dimension, which is convenient for subsequent processing.

[0154] S112. Perform missing value processing on several historical node information with the same dimension to obtain several complete historical node information; wherein the missing value processing is to obtain node information with missing competitiveness value in the several node information with the same dimension, input the node information with missing competitiveness value into the fuzzy logic system for processing to obtain the corresponding competitiveness value, and complete the information of the node with missing competitiveness value according to the obtained competitiveness value.

[0155] In this embodiment, the fuzzy logic system mainly includes four parts: a fuzzification processor, a fuzzy inference engine, a fuzzy rule base, and a defuzzification processor;

[0156] The fuzzification processor is used to fuzzify the input node information, transforming the input node information into membership values ​​in the fuzzy set of the fuzzy logic system. Commonly used membership functions include triangular membership function, trapezoidal membership function, Gaussian membership function and Sigmoid membership function.

[0157] The fuzzy inference engine is responsible for inferring and calculating the membership values ​​generated by the fuzzification processor based on the input node information according to the fuzzy rule base. It uses specific inference methods to transform the input fuzzy set into the output fuzzy set. Commonly used inference methods include the Mamdani fuzzy inference method and the Takagi-Sugeno fuzzy inference method.

[0158] The fuzzy rule base is used to store the knowledge base of fuzzy logic systems. The rules are in the form of "if-then" statements, describing the impact of different combinations of input variables on the output.

[0159] The function of a defuzzification processor is to transform the fuzzy set output by the fuzzy inference engine into specific precise values, thereby achieving the function of defuzzification. Commonly used methods include the center method, the maximum membership method, and the weighted average method.

[0160] S113. Perform data format conversion processing on several complete historical node information to obtain a historical node dataset.

[0161] In this embodiment, the data format conversion process refers to converting the data format into (X, Y) = (RE, DB, ND, CS, CV) based on the relationship between historical node information.

[0162] S120. Train and test the fuzzy width learning model based on the historical node dataset to obtain the trained fuzzy width learning model.

[0163] Specifically, the competitiveness value depends on an influencing factor, which includes one or more of the following: the relative remaining energy of the node (Energy_Re(i), the relative distance from the node to the base station (BS) (Dist_BS(i), the density of the node's neighbor nodes (Nei_Dens(i), and the historical cluster head state (CH_State(i)) of the node).

[0164] Specifically, the relative remaining energy Energy_Re(i) of the node is obtained by the following formula:

[0165]

[0166] Among them, Engrgy res (i) represents the remaining energy of node i, Engry init (i) represents the initial energy of the node;

[0167] Understandably, the more energy a node has remaining, the greater its capacity for data transmission, making it more suitable to serve as the cluster head node (CH) for data collection, and thus more competitive with other cluster head nodes.

[0168] And / or, the relative distance Dist_BS(i) from the node to the base station BS is obtained by the following formula:

[0169]

[0170] Where d1, d2, d3, and d4 represent the distances from the four boundary corners of the wireless sensor network to the base station BS, and max(.) is the maximum value function, representing the selection of the distance from d1, d2, d3, and d4 that is farthest from the base station BS; Dist BS (S i The distance () represents the Euclidean distance from node i to base station BS, which is obtained through the following calculation formula:

[0171]

[0172] Where x i y i Let x and y represent the x and y coordinates of node i, respectively. BS y BS These represent the horizontal and vertical coordinates of the base station BS, respectively.

[0173] Understandably, the closer a node is to the base station (BS), the less energy it needs for transmission, the more data forwarding tasks it can handle, and the greater its competitiveness as a cluster head node. While nodes at greater distances may have a wider coverage area, the energy required for data transmission also increases accordingly.

[0174] And / or, the neighbor node density Nei_Dens(i) of the node is specifically obtained by the following formula:

[0175]

[0176] Where, N w Let represent the total number of nodes in the wireless sensor network, and Neighbor(i) represent the number of neighboring nodes of node i, which is obtained through the following formula:

[0177]

[0178] Where j represents other nodes j in the wireless sensor network besides node i, and d ij R represents the distance between node i and the other nodes j. t This represents the maximum radius distance at which node i transmits relevant information, and δ(.) is an indicator function, where d ij ≤R t When δ = 1, δ = 0; otherwise, δ = 0.

[0179] It is understandable that a node i in radius R tThe node i sends an ADV message, including the sending node ID, within its range. When a neighboring node receives the broadcast message, it returns an ACK message, including the receiving node ID, node location, and remaining energy. After receiving the ACK message, node i stores the information in its adjacency list. The adjacency list of a node is updated in each round. The number of neighboring nodes of node i can be calculated from the adjacency list of node i. Understandably, the higher the density of neighboring nodes of a node, the better it is to collect and transmit data, and the lower the energy consumption for transmission and aggregation, making the node more suitable to serve as a cluster head node.

[0180] And / or, the historical cluster head state CH_State(i) of the node is obtained by the following formula:

[0181]

[0182] Where r represents the number of times node i is consecutively selected as the cluster head node;

[0183] Understandably, a node's repeated role as cluster head node leads to rapid energy consumption, and excessive consumption will cause the node to die, affecting the stable operation of the wireless sensor network. Therefore, it is necessary to record the historical status of nodes and replace cluster head nodes periodically. For nodes that have not served as cluster head nodes in the first two rounds, their original competitiveness should be maintained. For nodes that have served as cluster head nodes in the first two rounds, their competitiveness should be reduced. For nodes that have served as cluster head nodes in the first two rounds consecutively, their competitiveness should be significantly reduced.

[0184] Specifically, such as Figure 3 As shown, the fuzzy width learning model includes an input layer, a fuzzy feature layer, a feature enhancement layer, and an output layer; the fuzzy feature layer includes several fuzzy subsystems.

[0185] In this embodiment, the input layer inputs the preprocessed node information; the purpose of the fuzzy feature layer is to extract the feature information of the data samples, and the input data samples are fuzzy processed using the FCM fuzzy c-means clustering algorithm; the feature enhancement layer expands and enhances the feature information output by the fuzzy feature layer through nonlinear transformation to extract higher-dimensional feature representations; the calculation result of the output layer is the node's competitiveness value (CV), which serves as an indicator of a node's competitiveness in becoming a cluster head node;

[0186] The process of training and testing the fuzzy width learning system network model based on the historical node dataset includes the following steps:

[0187] S121. Divide the historical node dataset into a training set and a test set;

[0188] Preferably, the training set:test set ratio is set to 4:1;

[0189] S122. Set the number of fuzzy subsystems Q and the number of fuzzy rules K;

[0190] In this embodiment, the fuzzy feature layer contains Q fuzzy subsystems, where K represents the number of fuzzy rules in each fuzzy subsystem. Setting the values ​​of Q and K appropriately has a significant impact on the performance and training effect of the fuzzy width learning model.

[0191] S123. Initialize the coefficient weights of the fuzzy width learning model;

[0192] S124. The training set is input into the fuzzy width learning model through the input layer;

[0193] S125. In the fuzzy feature layer, an intermediate matrix Z is generated using several fuzzy subsystems of the fuzzy width learning model and the training set. n and feature node matrix F n , where n is the total number of training samples in the training set;

[0194] S126, the intermediate matrix Z n Each node feature value is input into the feature enhancement layer, and several enhanced nodes H are obtained through enhanced feature processing. m According to the feature node matrix F n and the aforementioned several enhancement nodes H m Obtain the target value Y, and quickly calculate the weight matrix W based on the target value Y and the pseudo-inverse.

[0195] S127. In the output layer, the preliminary parameters of the fuzzy width learning model are obtained according to the weight matrix W;

[0196] S128. Input the test set into the fuzzy width learning model and calculate the error, which includes the error rate ERR and the mean absolute percentage error MAPE.

[0197] S129. After optimizing the number of fuzzy subsystems Q and the number of fuzzy rules K based on the error rate ERR and the mean absolute percentage error MAPE, iteratively train the fuzzy width learning model, optimize the initial parameters to obtain the optimal model parameters, and complete the training of the fuzzy width learning model.

[0198] In this embodiment, the fuzzy width learning model is trained multiple times, and the values ​​of Q and K are searched within a certain range using a grid search in order to find the most suitable and optimal model.

[0199] Specifically, each training sample in the training set is N×M dimensional data, where N represents the number of training sample groups and M represents the feature information dimension of the training sample, specifically expressed as follows:

[0200] X = (x1, x2, ..., x s ,...,x N ) T ∈R N×M

[0201] Where x s =(x s1 ,x s2 ,...,x sm ,...,x sM ), s=1,2,…,N;

[0202] And / or, the setting of the number of fuzzy subsystems Q and the number of fuzzy rules K specifically includes:

[0203] In the q-th fuzzy subsystem, K is set q The fuzzy rules are represented in the following form:

[0204] If x s1 is and x s2 is … and x sM is

[0205] but

[0206] Where k = 1, 2, ..., K q x sM This represents one of the training samples in the training set. This represents a preset fuzzy set in the fuzzy subsystem. u is an adjustable parameter of the model rule. qg This represents the training sample g in the training set for the fuzzy rule K. q membership degree The intermediate matrix Z n The output of the q-th fuzzy subsystem, where u qg It can be represented as:

[0207]

[0208] Among them, c q c k This represents the cluster center corresponding to the fuzzy rule;

[0209] Then the training sample xs The output Z of the q-th fuzzy subsystem sq Represented as:

[0210]

[0211] in, The output Z for the q-th fuzzy subsystem sq single middle value The weights; then the intermediate matrix Z of the combined output of the Q fuzzy subsystems of the training set. n Represented as:

[0212] Z n =(Z1,Z2,…,Z n )

[0213] The deblurred output F of the blurry feature layer sq Represented as:

[0214]

[0215] in, These are adjustable parameters;

[0216] The defuzzification output F of the q-th fuzzy subsystem q for:

[0217]

[0218] in, diag{.} is a diagonal matrix;

[0219] The output F of the fuzzy feature layer n Represented as:

[0220]

[0221] And / or, the intermediate matrix Z n Each node feature value is input into the feature enhancement layer, and several enhanced nodes H are obtained through enhanced feature processing. m ,include:

[0222] The intermediate matrix Z n Z value of each feature node in I The value H mapped to the augment node J :

[0223]

[0224] Where, ξ J (.) is a mapping function. For random weights, For bias terms;

[0225] Then the several enhancement nodes H of the m features m The output is represented as:

[0226] H m =(H1,H2,…,H m )

[0227] And / or, according to the feature node matrix F n and the aforementioned several enhancement nodes H m To obtain the target value Y, the following steps are required:

[0228] The target value Y output by the output layer is:

[0229]

[0230] Among them, W f W represents the weight coefficients from the blurred feature layer to the output layer. h These are the weight coefficients from the feature enhancement layer to the output layer;

[0231] And / or, the weight matrix W is obtained by fast calculation based on the target value Y and the pseudo-inverse, specifically according to the following formula:

[0232] W=(Bω,H m ) + Y

[0233] Among them, (Bω,H m ) + =((Bω,H m ) T (Bω,H m )) -1 (Bω,H m ) T ;

[0234] And / or, the formula for calculating the Mean Absolute Percentage Error (MAPE) is expressed as follows:

[0235]

[0236] Among them, y v Let the v-th intermediate target true value be the target value Y. For y v The predicted value;

[0237] And / or, the formula for calculating the error rate ERR is expressed as:

[0238]

[0239] Among them, E baselineE represents the baseline error value of the fuzzy width learning model. model This represents the error value in the current training process of the fuzzy width learning model.

[0240] S130. Generate several nodes of the wireless sensor network and their corresponding node information. After preprocessing the several node information, input them into the trained fuzzy width learning model. Calculate the competitiveness value of the node corresponding to each node information through the fuzzy width learning model.

[0241] S140. Obtain the attributes of the wireless sensor network, and obtain the optimal number of cluster heads of the wireless sensor network based on the attributes of the wireless sensor network.

[0242] Specifically, the attributes of the wireless sensor network include the total number of nodes in the wireless sensor network, the size of the wireless sensor network, and the average distance from each node in the wireless sensor network to the base station (BS).

[0243] The step of obtaining the optimal number of cluster heads for the wireless sensor network based on its attributes includes:

[0244] Optimal number of cluster heads k opt Specifically, it can be obtained through the following formula:

[0245]

[0246] Where, N w Let d be the total number of nodes in the wireless sensor network, P be the size of the wireless sensor network, and d be the total number of nodes in the wireless sensor network. toBS The average distance from each node in the wireless sensor network to the base station (BS);

[0247] The average distance from each node in the wireless sensor network to the base station (BS) is specifically obtained using the following formula:

[0248]

[0249] In this embodiment, the optimal number of cluster heads refers to the most suitable number of cluster heads in the current wireless sensor network environment. Too many cluster heads will lead to additional communication consumption, while too few cluster heads will lead to excessive burden on the cluster heads and rapid energy consumption. To determine the optimal number of cluster heads, it is necessary to calculate based on the scale of the wireless sensor network, node density, and energy consumption.

[0250] S150. Based on the competitiveness value of each node, select a number of cluster head nodes, wherein the number of selected cluster head nodes is equal to the number of optimal cluster heads; nodes not selected as cluster head nodes are set as ordinary nodes.

[0251] Specifically, the step of selecting several cluster head nodes based on the competitiveness value of each node includes:

[0252] Sort all the nodes by competitiveness value from high to low, select the node with high competitiveness as the cluster head node, and the rest of the nodes as ordinary nodes.

[0253] And / or, the step of constructing several clusters in the wireless sensor network for data transmission based on several cluster head nodes and several ordinary nodes includes:

[0254] The clusters are constructed using a second cluster head strategy, a sub-cluster head strategy, and a minimum energy consumption criterion strategy.

[0255] The second cluster head strategy involves selecting the node with the highest competitiveness value (excluding the cluster head node) from the constructed cluster group to serve as the second cluster head, and the second cluster head assists the cluster head node in transmission work.

[0256] In this embodiment, the second cluster head strategy can effectively balance energy consumption within the cluster.

[0257] The secondary cluster head strategy involves selecting a node in the cluster that can replace the cluster head node as a secondary cluster head node. When the cluster head node fails or cannot work properly, the secondary cluster head node takes over as the cluster head node and completes the work of the cluster head node.

[0258] In this embodiment, the secondary cluster head strategy provides a fault-tolerant mechanism for wireless sensor networks, avoiding the impact of single-point cluster head failure on network operation;

[0259] The minimum energy consumption criterion strategy calculates the indirect energy consumption E of data transmission from the ordinary node to the base station BS via the cluster head node when the ordinary node transmits data in the wireless sensor network. idt And the direct energy consumption E of computing data being transmitted directly from the ordinary node to the base station BS. dt And compare the indirect energy consumption E using a cost function. idt and the direct energy consumption E dt The size of the path determines the path with the lowest energy consumption for transmission.

[0260] Specifically, the indirect energy consumption and / or direct energy consumption are affected by the following parameters:

[0261] The distance 'a' from a regular node to the cluster head node is specifically related to: E idt ∝a;

[0262] The distance b from the cluster head node to the base station BS is specifically related to: E idt ∝b;

[0263] The distance c from a regular node to a base station (BS) is specifically related to: E dt ∝c;

[0264] The indirect energy consumption E idt Specifically, it can be obtained through the following formula:

[0265] E idt =l·[(1+μ)E elec +∈ fs (a 2 +μb 2 )]

[0266] Where μ represents the preset data aggregation rate;

[0267] Among them, E elec The energy consumption coefficient of the path, ∈ fs This represents the power amplification factor of the existing free-space model;

[0268] The direct energy consumption E dt Specifically, it can be obtained through the following formula:

[0269] E dt =l·(E elec +∈ fs c 2 )

[0270] The indirect energy consumption E idt and the direct energy consumption E dt The energy consumption difference ΔE is obtained through the following formula:

[0271] △E=E idt -E dt =l·[μE elec +ε fs (a 2 +μb 2 -c 2 )]

[0272] When △E>0, it means E dt Numerically smaller, the ordinary node selects data to be transmitted directly from the ordinary node to the base station (BS) as the transmission path for data transmission; when ΔE < 0, it indicates that E idt Numerically smaller, the ordinary node data is indirectly transmitted from the ordinary node to the base station (BS) via the cluster head node as the transmission path for data transmission.

[0273] In this embodiment, the minimum energy consumption criterion strategy can reduce overall energy consumption by optimizing the transmission path, thereby further improving network performance and lifespan;

[0274] The parameter 'a' indicates that the farther a normal node is from the cluster head node, the greater the energy consumed by indirect data transmission through the cluster head. Therefore, non-cluster head nodes preferentially join clusters that are closer to them.

[0275] Parameter b indicates that the farther the cluster head node is from the base station (BS), the more energy the cluster head node needs to consume to transmit data packets to the base station (BS).

[0276] The parameter c indicates that the closer a normal node is to the base station (BS), the less energy it consumes to transmit data to the base station (BS), and direct transmission to the base station (BS) can be considered.

[0277] S160. Based on the cluster head nodes and the ordinary nodes, several clusters in the wireless sensor network are constructed for data transmission.

[0278] Specifically, the method further includes:

[0279] After the cluster is constructed, the base station (BS) sends an acknowledgment message (BS_info message) to each node in the wireless sensor network, including the ID of the cluster head node, the coordinates of the cluster head node, and the cluster to which the cluster head node belongs; each node starts working after receiving the acknowledgment message and enters the data collection and transmission phase.

[0280] During the data collection and transmission phase, the cluster head node performs data aggregation on the collected data; wherein, the data aggregation adopts the IA data aggregation model, and the cluster head node aggregates and forwards the received data packets according to the data aggregation rate, with the specific calculation formula as follows:

[0281] L agg =L r +μ×L r ×N p

[0282] Among them, L agg L represents the length of the aggregated data packet. r The length of the received data packet is represented by μ, and the data aggregation rate is represented by N. p This indicates the number of data packets received.

[0283] In this embodiment, the above operation takes into account that the data sensed by nearby nodes are similar or overlapping. The cluster head node performs data aggregation on the collected data, which can reduce the redundancy between data.

[0284] Based on the above methods and steps, this application completes the simulation experiment of Example 1, specifically including:

[0285] In Example 1, 5000 historical nodes and their corresponding information were collected to train and test the fuzzy width learning model. Subsequently, 200 sensor nodes were set up in a simulated wireless sensor network environment, randomly distributed in a 100m*100m scene. Each node initially had the same amount of initial energy of 0.5J, and its position could not be moved after deployment, and its energy could not be replenished. Preferably, as shown in Table 1 of the fuzzy logic rules for competing cluster head nodes in this method.

[0286] Table 1. Fuzzy logic rules for competing cluster head nodes.

[0287]

[0288]

[0289] If a node has more remaining energy, is closer to the base station (BS), has a higher density of surrounding nodes, and does not have a history of serving as a cluster head multiple times consecutively, then the node's competitiveness as a cluster head will be correspondingly greater.

[0290] In the node competitiveness calculation, feature information derived from node information is used as an influencing factor, and a fuzzy width learning model is adopted to obtain the competitiveness value of a node serving as a cluster head node.

[0291] Table 2 shows the cluster head competitiveness values ​​calculated based on the fuzzy width learning model in the simulation experiment of Example 1.

[0292] Table 2 calculates the cluster head competitiveness values ​​of the nodes.

[0293]

[0294] Preferably, in this embodiment 1, the optimal number of cluster heads is 14, which can be determined from the simulation network environment size of 100m*100m.

[0295] In this embodiment 1, the optimal number of cluster heads is 14. Due to the second cluster head strategy, only 7 clusters are generated, and each cluster contains 1 cluster head and 1 second cluster head.

[0296] This application also includes Embodiment 2, which, compared to Embodiment 1, conducts simulation experiments in a network environment with randomly distributed energy;

[0297] This application also includes Embodiment 3, which, compared to Embodiment 1, conducts simulation experiments in a network environment with a small initial energy.

[0298] This application also includes Embodiment 4, which, compared to Embodiment 1, conducts a simulation experiment in a network environment with mobile nodes;

[0299] Examples 2 and 3 differ from the simulation parameter configuration of the Base Scenario, showcasing different performances under various scenarios: one with uneven energy distribution and the other with a small initial energy. The experiments utilize Matlab simulation parameters to model the process of a real wireless sensor network environment, collecting all node information and inputting it into a fuzzy width learning model to optimize the entire clustering process. To compare and analyze the applicability under dynamic node conditions, Example 4 conducts a simulation experiment under the condition of moving nodes.

[0300] Figure 4 The average remaining energy performance of nodes in different embodiments of the present invention and comparative method are shown. Figure 5 The average remaining energy performance fluctuations of nodes in different embodiments of each method are shown, where FBLSC represents the present invention. Experimental results show that, compared with other existing methods, the present invention exhibits better performance fluctuations under different environments and has stronger adaptability. Compared with the huge time overhead generated by complex methods, the present invention can significantly reduce the training time of the model through the rapid training capability of the fuzzy width learning model, while maintaining good accuracy. At the same time, it has incremental characteristics and shows good scalability.

[0301] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the technical solution of the present invention, and are not intended to limit the specific implementation of the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the claims of the present invention should be included within the protection scope of the claims of the present invention.

Claims

1. A method for energy-efficient competitive clustering in wireless sensor networks based on fuzzy width learning, characterized in that, The method includes: Collect information on several historical nodes of a wireless sensor network, and preprocess the information on several historical nodes to generate a historical node dataset. The fuzzy width learning model is trained and tested based on the historical node dataset to obtain a trained fuzzy width learning model. Generate several nodes of the wireless sensor network and their corresponding node information. After preprocessing the node information, input it into the trained fuzzy width learning model. Calculate the competitiveness value of the node corresponding to each node information through the fuzzy width learning model. Obtain the attributes of the wireless sensor network, and obtain the optimal number of cluster heads for the wireless sensor network based on the attributes of the wireless sensor network. Based on the competitiveness value of each node, a number of cluster head nodes are selected, wherein the number of selected cluster head nodes is equal to the number of optimal cluster heads; nodes not selected as cluster head nodes are set as ordinary nodes. Based on the cluster head nodes and the ordinary nodes, several clusters are constructed in the wireless sensor network for data transmission.

2. The energy efficiency competitive clustering method for wireless sensor networks based on fuzzy width learning according to claim 1, characterized in that, The step of preprocessing the historical node information to generate a historical node dataset includes: Data normalization processing is performed on several historical node information to obtain several historical node information with the same dimension. Missing value processing is performed on several historical node information with the same dimension to obtain several complete historical node information; wherein the missing value processing is to obtain node information with missing competitiveness value in the several node information with the same dimension, input the node information with missing competitiveness value into a fuzzy logic system for processing to obtain the corresponding competitiveness value, and fill in the information of the node with missing competitiveness value according to the obtained competitiveness value. The historical node dataset is obtained by converting the data format of several complete historical node information.

3. The energy efficiency competitive clustering method for wireless sensor networks based on fuzzy width learning according to claim 1, characterized in that, The competitiveness value depends on an influencing factor, which includes one or more of the following: the relative remaining energy of the node, Energy_Re(i), the relative distance of the node to the base station, Dist_BS(i), the density of the node's neighbor nodes, Nei_Dens(i), and the historical cluster head state of the node, CH_State(i).

4. The energy efficiency competitive clustering method for wireless sensor networks based on fuzzy width learning according to claim 3, characterized in that, The relative remaining energy Energy_Re(i) of the node is obtained by the following formula: Energy res (i) represents the remaining energy of node i. init (i) represents the initial energy of node i; And / or, The relative distance Dist_BS(i) from the node to the base station BS is obtained by the following formula: Where d1, d2, d3, and d4 represent the distances from the four boundary corners of the wireless sensor network scenario to the base station BS, and max(.) is the maximum value function, representing the selection of the distance from d1, d2, d3, and d4 that is farthest from the base station BS; Dist BS (S i The distance () represents the Euclidean distance from node i to the base station BS, which is obtained through the following calculation formula: Where x i y i Let x and y represent the x and y coordinates of node i, respectively. BS y BS These represent the horizontal and vertical coordinates of the base station BS, respectively. And / or, the neighbor node density Nei_Dens(i) of node i is specifically obtained by the following formula: Where, N w The total number of nodes in the wireless sensor network is represented by ; Neighbor(i) represents the number of neighboring nodes of node i, which is obtained by the following formula: Where j represents other nodes j in the wireless sensor network besides node i, and d ij R represents the distance between node i and the other nodes j. t This represents the maximum radius distance at which node i transmits relevant information, and δ(.) is an indicator function, where d ij ≤R t When δ = 1, δ = 0; otherwise, δ = 0. And / or, the historical cluster head state CH_State(i) of the node is obtained by the following formula: Where r represents the number of times node i is consecutively selected as the cluster head node.

5. The energy efficiency competitive clustering method for wireless sensor networks based on fuzzy width learning according to claim 1, characterized in that, The fuzzy width learning model includes an input layer, a fuzzy feature layer, a feature enhancement layer, and an output layer; the fuzzy feature layer includes several fuzzy subsystems. The step of training and testing the fuzzy width learning system network model based on the historical node dataset includes: The historical node dataset is divided into a training set and a test set; Define the number of fuzzy subsystems Q and the number of fuzzy rules K; Initialize the coefficient weights of the fuzzy width learning model; The training set is input into the fuzzy width learning model through the input layer; In the fuzzy feature layer, an intermediate matrix Z is generated through several fuzzy subsystems of the fuzzy width learning model and the training set. n and feature node matrix F n , where n is the total number of training samples in the training set; The intermediate matrix Z n Each node feature value is input into the feature enhancement layer, and several enhanced nodes H are obtained through enhanced feature processing. m According to the feature node matrix F n and the aforementioned several enhancement nodes H m Obtain the target value Y, and quickly calculate the weight matrix W based on the target value Y and the pseudo-inverse. In the output layer, the preliminary parameters of the fuzzy width learning model are obtained based on the weight matrix W; The test set is input into the fuzzy width learning model to calculate the error, which includes the error rate ERR and the mean absolute percentage error MAPE. After optimizing the number of fuzzy subsystems Q and the number of fuzzy rules K based on the error rate ERR and mean absolute percentage error MAPE, the fuzzy width learning model is iteratively trained. The initial parameters are optimized to obtain the optimal model parameters, and the training of the fuzzy width learning model is completed.

6. The energy efficiency competitive clustering method for wireless sensor networks based on fuzzy width learning according to claim 5, characterized in that, Each training sample in the training set is N×M dimensional data, where N represents the number of training sample groups and M represents the feature information dimension of the training sample, specifically as follows: X=(x1,x2,…,x s ,…,x N ) T ∈R N×M Where x s =(x s1 ,x s2 ,...,x sm ,...,x sM ), s=1,2,…,N; And / or, the setting of the number of fuzzy subsystems Q and the number of fuzzy rules K specifically includes: In the q-th fuzzy subsystem, K is set q The fuzzy rules are represented in the following form: if but Where k = 1, 2, ..., K q x sm This represents one of the training samples in the training set. This represents a preset fuzzy set in the fuzzy subsystem. u is an adjustable parameter of the model rule. qg This represents the training sample g in the training set for the fuzzy rule K. q membership degree The intermediate matrix Z n The output of the q-th fuzzy subsystem, where u qg It can be represented as: Among them, c q c k This represents the cluster center corresponding to the fuzzy rule; Then the training sample x s The output Z of the q-th fuzzy subsystem sq Represented as: in, The output Z for the q-th fuzzy subsystem sq single middle value The weights; The intermediate matrix Z of the training set output by the combination of the Q fuzzy subsystems n Represented as: Z n =(Z1,Z2,…,Z n The deblurred output F of the blurred feature layer sq Represented as: in, These are adjustable parameters; The defuzzification output F of the q-th fuzzy subsystem q for: in, diag{.} is a diagonal matrix; The output F of the fuzzy feature layer n Represented as: And / or, the intermediate matrix Z n Each node feature value is input into the feature enhancement layer, and several enhanced nodes H are obtained through enhanced feature processing. m ,include: The intermediate matrix Z n Z value of each feature node in I The value H mapped to the augment node J : Where, ξ J (.) is a mapping function. For random weights, For bias terms; Then the several enhancement nodes H of the m features m The output is represented as: H m =(H1,H2,…,H m ) And / or, according to the feature node matrix F n and the aforementioned several enhancement nodes H m To obtain the target value Y, the following steps are required: The target value Y output by the output layer is: Among them, W f W represents the weight coefficients from the blurred feature layer to the output layer. h These are the weight coefficients from the feature enhancement layer to the output layer; And / or, the weight matrix W is obtained by fast calculation based on the target value Y and the pseudo-inverse, specifically according to the following formula: W=(Bω,H m ) + Y Among them, (Bω, H m ) + =((Bω,H m ) T (Bω,H m )) -1 (Bω,H m ) T ; And / or, the formula for calculating the Mean Absolute Percentage Error (MAPE) is expressed as follows: Among them, y v Let the v-th intermediate target true value be the target value Y. For y v The predicted value; And / or, the formula for calculating the error rate ERR is expressed as: Among them, E baseline E represents the baseline error value of the fuzzy width learning model. model This represents the error value in the current training process of the fuzzy width learning model.

7. The energy efficiency competitive clustering method for wireless sensor networks based on fuzzy width learning according to claim 1, characterized in that, The attributes of the wireless sensor network include the total number of nodes in the wireless sensor network, the size of the wireless sensor network, and the average distance from each node in the wireless sensor network to the base station (BS). The step of obtaining the optimal number of cluster heads for the wireless sensor network based on its attributes includes: Optimal number of cluster heads k opt Specifically, it can be obtained through the following formula: Where, N w Let d be the total number of nodes in the wireless sensor network, P be the size of the wireless sensor network, and d be the total number of nodes in the wireless sensor network. toBS The average distance from each node in the wireless sensor network to the base station (BS); The average distance from each node in the wireless sensor network to the base station (BS) is specifically obtained using the following formula:

8. A method for energy-efficient competitive clustering of wireless sensor networks based on fuzzy width learning according to any one of claims 1 to 7, characterized in that, The step of selecting several cluster head nodes based on the competitiveness value of each node includes: Sort all the nodes by competitiveness value from high to low, select the node with high competitiveness as the cluster head node, and the rest of the nodes as ordinary nodes. And / or, The step of constructing several clusters in the wireless sensor network based on several cluster head nodes and several ordinary nodes for data transmission includes: The clusters are constructed using a second cluster head strategy, a sub-cluster head strategy, and a minimum energy consumption criterion strategy. The second cluster head strategy involves selecting the node with the highest competitiveness value (excluding the cluster head node) from the constructed cluster group to serve as the second cluster head, and the second cluster head assists the cluster head node in transmission work. The secondary cluster head strategy involves selecting a node in the cluster that can replace the cluster head node as a secondary cluster head node. When the cluster head node fails or cannot work properly, the secondary cluster head node takes over as the cluster head node and completes the work of the cluster head node. The minimum energy consumption criterion strategy calculates the indirect energy consumption E of data transmission from the ordinary node to the base station BS via the cluster head node when the ordinary node transmits data in the wireless sensor network. idt And the direct energy consumption E of computing data being transmitted directly from the ordinary node to the base station BS. dt And compare the indirect energy consumption E using a cost function. idt and the direct energy consumption E dt The size of the path determines the path with the lowest energy consumption for transmission.

9. The energy efficiency competitive clustering method for wireless sensor networks based on fuzzy width learning according to claim 8, characterized in that, The indirect and / or direct energy consumption is affected by the following parameters: The distance 'a' from a regular node to the cluster head node is specifically related to: E idt ∝a; The distance b from the cluster head node to the base station BS is specifically related to: E idt ∝b; The distance c from a regular node to a base station (BS) is specifically related to: E dt ∝c; The indirect energy consumption E idt Specifically, it can be obtained through the following formula: And idt =l·[(1+μ)E elec +∈ fs (to 2 +μb 2 )] Where μ represents the preset data aggregation rate; Among them, E elec The energy consumption coefficient of the path, ∈ fs This represents the power amplification factor of the existing free-space model; The direct energy consumption E dt Specifically, it can be obtained through the following formula: AND dt =l·(E elec +∈ fs c 2 ) The indirect energy consumption E idt and the direct energy consumption E dt The energy consumption difference ΔE is obtained through the following formula: △E=E idt -E dt =l·[μE elec +e fs (a 2 +μb 2 -c 2 )] When △E>0, it means E dt Numerically smaller, the ordinary node selects data to be transmitted directly from the ordinary node to the base station (BS) as the transmission path for data transmission; when ΔE < 0, it indicates that E idt Numerically smaller, the ordinary node data is indirectly transmitted from the ordinary node to the base station (BS) via the cluster head node as the transmission path for data transmission.

10. A method for energy-efficient competitive clustering of wireless sensor networks based on fuzzy width learning according to any one of claims 1 to 7, characterized in that, The method further includes: After the cluster is constructed, the base station (BS) sends an acknowledgment message (BS_info message) to each node in the wireless sensor network, including the ID of the cluster head node, the coordinates of the cluster head node, and the cluster to which the cluster head node belongs; each node starts working after receiving the acknowledgment message and enters the data collection and transmission phase. During the data collection and transmission phase, the cluster head node performs data aggregation on the collected data; wherein, the data aggregation adopts the IA data aggregation model, and the cluster head node aggregates and forwards the received data packets according to the data aggregation rate, with the specific calculation formula as follows: L agg =L r +μ×L r ×N p Among them, L agg L represents the length of the aggregated data packet. r The length of the received data packet is represented by μ, and the data aggregation rate is represented by N. p This indicates the number of data packets received.

Images