Node configuration optimization method for wireless sensor networks based on VGPSO algorithm

Abstract

This application provides a wireless sensor network node configuration optimization method based on the VGPSO algorithm, which relates to the field of wireless sensor networks. The method includes: defining wireless sensor network node configuration optimization as problem M; guiding an improved particle swarm algorithm using a random Halton sequence, a multi-factor cosine dynamic parameter adjustment strategy, a multi-population multi-strategy evolutionary approach, and a Voronoi diagram; solving problem M using the improved particle swarm algorithm and outputting the global optimal position of each sensor node in the wireless sensor network. A multi-factor cosine dynamic parameter adjustment strategy is introduced to balance global and local optimization search capabilities and enhance the diversity of the particle population. Voronoi polygons are used to guide the evolutionary direction, and the direction of coverage holes is found by calculating the Voronoi centroid to optimize coverage holes during population evolution.

Description

Technical Field

[0001] The present application relates to the field of wireless sensor networks, and in particular to a method for optimizing node configuration in wireless sensor networks based on a VGPSO algorithm. Background Art

[0002] In the 5G era, the rapid development of IoT technology has driven the widespread application of wireless sensor networks (WSNs) in many industrial fields, such as power equipment monitoring and ship communications. WSNs are composed of numerous sensor nodes, which generally possess computing, storage, and wireless communication capabilities. WSN performance is often measured by sensor node coverage. To avoid random deployment and waste of sensor resources while reducing network construction costs, optimizing the configuration of sensor nodes in wireless sensor networks is crucial. This optimized configuration achieves a more even distribution of sensor nodes and wider coverage, effectively utilizing spatial resources and improving the efficiency of environmental perception and information acquisition. This not only improves network performance and reliability but also reduces network construction costs. Summary of the Invention

[0003] The purpose of the present invention is to provide a wireless sensor network node configuration optimization method based on the VGPSO algorithm in order to solve the problem of low coverage of existing sensor node configuration methods.

[0004] The above-mentioned purpose of this application is achieved through the following technical solutions:

[0005] S1: Define wireless sensor network node configuration optimization as a problem ;

[0006] S2: Improve the particle swarm optimization algorithm through random Halton sequence, multi-factor cosine dynamic parameter adjustment strategy, multi-population and multi-strategy evolution method and Voronoi diagram guidance;

[0007] S3: The problem is solved by the improved particle swarm optimization algorithm Solve and output the global optimal position of each sensor node in the wireless sensor network.

[0008] Optionally, step S3 includes:

[0009] S31: Setup Issues The basic parameters and parameters of improved particle swarm algorithm;

[0010] S32: Determine the optimal number of sensor nodes in the wireless sensor network;

[0011] S33: Initialize the population using random Halton sequence;

[0012] S34: Based on the cosine function, a control factor is introduced to design a multi-factor cosine dynamic parameter adjustment strategy; through the multi-factor cosine dynamic parameter adjustment strategy, the inertia weight and the global and individual learning factors are dynamically adjusted;

[0013] S35: Using a multi-population and multi-strategy evolutionary approach, the initial population is divided into three sub-populations according to a preset ratio based on the fitness function values ​​of the individuals in the population; using a multi-population and multi-strategy evolutionary approach, a speed update strategy is designed for each sub-population and the sub-populations are updated;

[0014] S36: Perform reflection out-of-bounds processing and position update on the updated sub-population;

[0015] S37: Calculate the boot probability, Centroid and guide step length;

[0016] S38: Update the position of the sensor node and the current individual fitness value through the guidance probability and guidance step size; compare the current individual fitness value of the population with the optimal fitness value of the historical evolution generation, update the optimal solution of the position of the sensor node until the maximum evolution generation is reached, and output the global optimal position of each sensor node.

[0017] Optionally, step S31 includes:

[0018] Defining the Problem The fitness function ,question The decision vector , Dimension ,as well as The length of the upper bound vector of the search region in the problem and the lower boundary vector , and the width upper boundary vector and the lower boundary vector ;

[0019] Set the sensor's sensing radius to , number of nodes , the discrete granularity is 1;

[0020] Determine the relevant parameters of the particle swarm algorithm, including: evolutionary generations Maximum speed , minimum speed , initialize the population size , control factor ;

[0021] The problem M The search area in is discretized into pixels, if the wireless sensor network covers pixels, the coverage for .

[0022] Optionally, step S32 includes:

[0023] set up and They represent the locations of three sensor nodes in the wireless sensor network; and The triangle formed by three points in a two-dimensional plane is an equilateral triangle; let A represent the center point of the equilateral triangle;

[0024] Line segment is the radius , and As the center of the circle, construct three circles, point A is the intersection of the three circles, point B is the center of the circle Heyuan The intersection of

[0025] Connecting line segments and line segments , the intersection of the two line segments is point C;

[0026] According to the properties of the circle , calculate the optimal number of sensor nodes , the formula is as follows:

[0027]

[0028] in Represents a line segment length; Represents a line segment and line segments Angle; Represents a line segment length.

[0029] Optionally, step S33 includes:

[0030] exist Within the scope, use Sequence initialization random generation Individuals, that is The solution vector of each individual is , where the odd-numbered bits range from ], the range of even-numbered digits is ;

[0031] individual Together they constitute the parent population, and the formula for initializing the parent population is as follows:

[0032]

[0033] is The range of sequence generation values ​​is A randomly distributed variable, The formula for the process is as follows:

[0034]

[0035] in and Represent the upper and lower boundaries of odd-numbered digits, respectively. and Respectively represent the upper and lower boundaries of even digits; Indicates the first Individual odd-dimensional variables, Indicates the first Even-dimensional variables of individuals, and constitutes an individual ; To generate the The prime number of the cardinality of the dimension, Represents an integer In the cardinality The next number of digits; is the cardinality The next No. Dimensional Sequence value, is an integer In the cardinality The index of the highest digit under ; is an integer In the cardinality The next digits.

[0036] Optionally, step S34 includes:

[0037] Introducing control factors , controlled by the factor and the current generation of the particle swarm algorithm calculate ;

[0038] Use the cosine function to adjust the inertia weight , the formula is as follows:

[0039]

[0040] Using inertia weight and , calculate the global and individual learning factors and , the formula is as follows:

[0041]

[0042] in ; Indicates an intermediate parameter.

[0043] Optionally, step S35 includes:

[0044] According to the fitness function value of the individuals in the population, the initial population is divided into three sub-populations according to the preset ratio , according to the fitness level, they are: development sub-population, moderate sub-population and exploration sub-population;

[0045] Speed ​​update strategies include: elite learning strategy, coordinated learning strategy and generalized learning strategy;

[0046] Development subpopulation Use elite learning strategy, moderate subpopulation Use coordinated learning strategies to explore subpopulations Use generalized learning strategies;

[0047] The speed update formula of the elite learning strategy is as follows:

[0048]

[0049] The speed update formula of the coordinated learning strategy is as follows:

[0050]

[0051] The speed update formula of the generalized learning strategy is as follows:

[0052]

[0053] in is a random integer, For the The speed of the next generation of individuals, For the current The speed of an individual, and is an individual randomly selected from the parent population, It is one of the individuals in the population whose fitness function value ranks in the top preset percentage from small to large; is the current individual historical optimal solution; is a random number in the range of [-0.5,0.5], is the value range of each variable; and The random value range is [0,0.5].

[0054] Optionally, step S36 includes:

[0055] If the speed If the speed exceeds the specified range, the reflection method with a certain interval is used to reflect the variable value back to the legal range. The formula of the reflection process is as follows:

[0056]

[0057] in The function is used to find the remainder of non-integer numbers. Indicates the minimum speed; Indicates the maximum speed;

[0058] Update the position of the sensor node: according to the speed Calculate the position of the next generation of population individuals , for position Perform reflection out-of-bounds processing.

[0059] Optionally, step S37 includes:

[0060] The formula for calculating the bootstrap probability is as follows:

[0061]

[0062] in , used to control the steepness of the logistic function, represents the threshold that controls the maximum increase in the bootstrap probability, For continuous algebra without improvement; is the minimum bootstrap probability, is the maximum bootstrap probability; Indicates the Bootstrap probability of generation;

[0063] According to the population position of the current generation, split Axis direction coordinate list and Y axis direction coordinate list, get the circle center list ; Calculate based on the circle center list centroid , the formula is as follows:

[0064]

[0065] in is the center of the circle list The X-axis coordinate of the coordinate, is the center of the circle list The Y-axis coordinate of the coordinate; Indicates the current The number of vertices of the polygon;

[0066] Traverse the list of circle centers obtained by the population layout position, In the list, find the circle with the current center The distance is minimum and smaller than the perception radius The guiding centroid );

[0067] The formula for calculating the bootstrap step size is as follows:

[0068]

[0069] in is the current center and centroid distance; The center of the circle and The angle difference of the centroid, and are the guiding step length and guiding angle, is the guide vector; is a randomization function, cos and sin are trigonometric functions; is the function for calculating the angle.

[0070] Optionally, step S38 includes:

[0071] The formula for updating the position of the sensor node is as follows:

[0072]

[0073] in, is the bootstrap probability; is the randomization function.

[0074] The beneficial effects of the technical solution provided by this application are:

[0075] 1. When solving the problem of node configuration optimization in wireless sensor networks, the initialization population obtained by random Halton sequence initialization greatly speeds up the early search process of the algorithm.

[0076] 2. Multi-factor cosine dynamic parameter adjustment strategy: Based on the existing dynamic adjustment of inertia weight and learning factor parameters according to the cosine function, in order to enhance the diversity of the particle population, a new control factor is introduced to balance the global and local optimization search capabilities:

[0077] 3. Multi-population, multi-strategy evolutionary approach: Based on the fitness function values ​​of individual populations, individuals are divided into exploitation sub-populations, moderated sub-populations, and exploration sub-populations. The exploitation sub-population focuses on improving the population's utilization capacity; the moderated sub-population randomly explores or exploits to achieve a balance between exploration and utilization; and the exploration sub-population focuses on improving the population's exploration capacity. The mutation strategy of the differential evolution algorithm is introduced during the speed update process to design a speed update strategy for each sub-population.

[0078] 4. Voronoi polygons guide evolutionary direction: Optimize coverage holes that appear during population evolution. Use Voronoi polygons to find the direction of coverage holes and calculate the Voronoi centroid. Through the guidance process, the algorithm guides the further evolutionary direction of some individuals, improving the local optimization ability of the algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] The present application will be further described below with reference to the accompanying drawings and embodiments, in which:

[0080] Figure 1 is a schematic diagram of sensor coverage in an embodiment of the present application;

[0081] Figure 2 Schematic diagram of the boot area in the embodiment of the present application;

[0082] Figure 3 is a randomly initialized node coverage graph in an embodiment of the present application;

[0083] Figure 4 It is a node coverage graph initialized based on Halton sequence in an embodiment of the present application;

[0084] Figure 5 This is a VGPSO coverage simulation diagram in an embodiment of the present application;

[0085] Figure 6 It is a graph showing the coverage change of various algorithms in the embodiments of the present application;

[0086] Figure 7 It is a flow chart in an embodiment of the present application. DETAILED DESCRIPTION

[0087] In order to have a clearer understanding of the technical features, purposes and effects of this application, the specific implementation methods of this application are now described in detail with reference to the accompanying drawings.

[0088] The embodiments of the present application provide a method for optimizing node configuration in a wireless sensor network based on a VGPSO algorithm.

[0089] Please refer to Figure 7 , Figure 7Flowchart of a wireless sensor network node configuration optimization method based on the VGPSO algorithm in an embodiment of the present application, including:

[0090] S1: Define wireless sensor network node configuration optimization as a problem ;

[0091] S2: Improve the particle swarm optimization algorithm through random Halton sequence, multi-factor cosine dynamic parameter adjustment strategy, multi-population and multi-strategy evolution method and Voronoi diagram guidance;

[0092] S3: The problem is solved by the improved particle swarm optimization algorithm Solve and output the global optimal position of each sensor node in the wireless sensor network.

[0093] Step S3 includes:

[0094] S31: Setup Issues The basic parameters and parameters of improved particle swarm algorithm;

[0095] Step S31 includes:

[0096] Defining the Problem The fitness function ,question The decision vector , Dimension ,as well as The length of the upper bound vector of the search region in the problem and the lower boundary vector , and the width upper boundary vector and the lower boundary vector ;

[0097] Set the sensor's sensing radius to , number of nodes , the discrete granularity is 1;

[0098] Determine the relevant parameters of the particle swarm algorithm, including: evolutionary generations Maximum speed , minimum speed , initialize the population size , control factor ;

[0099] The problem M The search area in is discretized into pixels, if the wireless sensor network covers pixels, the coverage for .

[0100] S32: Determine the optimal number of sensor nodes in the wireless sensor network;

[0101] Step S32 includes:

[0102] set up and They represent the locations of three sensor nodes in the wireless sensor network; and The triangle formed by three points in a two-dimensional plane is an equilateral triangle; let A represent the center point of the equilateral triangle;

[0103] Line segment is the radius , and As the center of the circle, construct three circles, point A is the intersection of the three circles, point B is the center of the circle Heyuan The intersection of

[0104] Connecting line segments and line segments , the intersection of the two line segments is point C;

[0105] According to the properties of the circle , calculate the optimal number of sensor nodes , the formula is as follows:

[0106]

[0107] in Represents a line segment length; Represents a line segment and line segments Angle; Represents a line segment length.

[0108] Specifically, the triangle formed by the sensor nodes is an equilateral triangle, such as Figure 1 shown.

[0109] S33: Initialize the population using random Halton sequence;

[0110] Step S33 includes:

[0111] exist Within the scope, use Sequence initialization random generation Individuals, that is The solution vector of each individual is , where the odd-numbered bits range from ], the range of even-numbered digits is ;

[0112] individual Together they constitute the parent population, and the formula for initializing the parent population is as follows:

[0113]

[0114] is The range of sequence generation values ​​is A randomly distributed variable, The formula for the process is as follows:

[0115]

[0116] in and Represent the upper and lower boundaries of odd-numbered digits, respectively. and Respectively represent the upper and lower boundaries of even digits; Indicates the first Individual odd-dimensional variables, Indicates the first Even-dimensional variables of individuals, and constitutes an individual ; To generate the The prime number of the cardinality of the dimension, Represents an integer In the cardinality The next number of digits; is the cardinality The next No. Dimensional Sequence value, is an integer In the cardinality The index of the highest digit under ; is an integer In the cardinality The next Number of digits.

[0117] S34: Based on the cosine function, a control factor is introduced to design a multi-factor cosine dynamic parameter adjustment strategy; through the multi-factor cosine dynamic parameter adjustment strategy, the inertia weight and the global and individual learning factors are dynamically adjusted;

[0118] Step S34 includes:

[0119] Introducing control factors , controlled by the factor and the current generation of the particle swarm algorithm calculate ;

[0120] Use the cosine function to adjust the inertia weight , the formula is as follows:

[0121]

[0122] Using inertia weight and , calculate the global and individual learning factors and , the formula is as follows:

[0123]

[0124] in ; Indicates an intermediate parameter.

[0125] Specifically, through adaptive adjustment strategies, the exploration and exploitation capabilities of particles are dynamically balanced during the optimization process, thereby improving the overall performance of the algorithm.

[0126] S35: Using a multi-population and multi-strategy evolutionary approach, the initial population is divided into three sub-populations according to a preset ratio based on the fitness function values ​​of the individuals in the population; using a multi-population and multi-strategy evolutionary approach, a speed update strategy is designed for each sub-population and the sub-populations are updated;

[0127] Step S35 includes:

[0128] According to the fitness function value of the individuals in the population, the initial population is divided into three sub-populations according to the preset ratio , according to the fitness level, they are: development sub-population, moderate sub-population and exploration sub-population;

[0129] Speed ​​update strategies include: elite learning strategy, coordinated learning strategy and generalized learning strategy;

[0130] Development subpopulation Use elite learning strategy, moderate subpopulation Use coordinated learning strategies to explore subpopulations Use generalized learning strategies;

[0131] The speed update formula of the elite learning strategy is as follows:

[0132]

[0133] The speed update formula of the coordinated learning strategy is as follows:

[0134]

[0135] The speed update formula of the generalized learning strategy is as follows:

[0136]

[0137] in is a random integer, For the The speed of the next generation of individuals, For the current The speed of an individual, and is an individual randomly selected from the parent population, It is one of the individuals in the population whose fitness function value ranks in the top preset percentage from small to large; is the current individual historical optimal solution; is a random number in the range of [-0.5,0.5], is the value range of each variable; and The random value range is [0,0.5].

[0138] S36: Perform reflection out-of-bounds processing and position update on the updated sub-population;

[0139] Step S36 includes:

[0140] If the speed If the speed exceeds the specified range, the reflection method with a certain interval is used to reflect the variable value back to the legal range. The formula of the reflection process is as follows:

[0141]

[0142] in The function is used to find the remainder of non-integer numbers. Indicates the minimum speed; Indicates the maximum speed;

[0143] Update the position of the sensor node: according to the speed Calculate the position of the next generation of population individuals , for position Perform reflection out-of-bounds processing.

[0144] S37: Calculate the boot probability, Centroid and guide step length;

[0145] Step S37 includes:

[0146] The formula for calculating the bootstrap probability is as follows:

[0147]

[0148] in , used to control the steepness of the logistic function, represents the threshold that controls the maximum increase in the bootstrap probability, For continuous algebra without improvement; is the minimum bootstrap probability, is the maximum bootstrap probability; Indicates the Bootstrap probability of generation;

[0149] According to the population position of the current generation, split Axis direction coordinate list and Y axis direction coordinate list, get the circle center list ; Calculate based on the circle center list centroid , the formula is as follows:

[0150]

[0151] in is the center of the circle list The X-axis coordinate of the coordinate, is the center of the circle list The Y-axis coordinate of the coordinate; Indicates the current The number of vertices of the polygon;

[0152] Traverse the list of circle centers obtained from the population layout position, In the list, find the circle with the current center The distance is minimum and smaller than the perception radius The guiding centroid );

[0153] The formula for calculating the bootstrap step size is as follows:

[0154]

[0155] in is the current center and centroid distance; The center of the circle and The angle difference of the centroid, and are the guiding step length and guiding angle, is the guide vector; is a randomization function, cos and sin are trigonometric functions; is the function for calculating the angle.

[0156] Specifically, Figure 2 A schematic diagram of the boot area.

[0157] S38: Update the position of the sensor node and the current individual fitness value through the guidance probability and guidance step size; compare the current individual fitness value of the population with the optimal fitness value of the historical evolution generation, update the optimal solution of the position of the sensor node until the maximum evolution generation is reached, and output the global optimal position of each sensor node.

[0158] Step S38 includes:

[0159] The formula for updating the position of the sensor node is as follows:

[0160]

[0161] in, is the bootstrap probability; is the randomization function.

[0162] This patent conducts simulation experiments and performance tests on the proposed method on the MATLAB R2022b platform. To ensure experimental fairness, all algorithms are set with the same parameters: the search area is 50 m × 50 m, the number of nodes is the theoretical number of nodes (calculated to be 42), the population size is 100, the perception radius is 5 m, and the number of iterations is set to 500. The WSN coverage of the improved particle swarm optimization algorithm guided by the Voronoi diagram (VGPSO) is compared with that of the particle swarm optimization algorithm (PSO), chaotic particle swarm optimization algorithm (CPSO), fuzzy particle swarm optimization algorithm (FPSO), unified particle swarm optimization algorithm (UPSO), and extended particle swarm optimization algorithm (XPSO).

[0163] Through Figure 3 and Figure 4 Comparative analysis shows that compared with random deployment of sensor nodes, the deployment method based on random Halton sequence makes the distribution of sensor nodes more uniform in the initialization phase, thereby improving the initial coverage.

[0164] Figure 5 This is the coverage simulation diagram based on the improved particle swarm optimization algorithm (VGPSO) of this application.

[0165] from Figure 6The coverage curves of six algorithms used in the WSN coverage optimization process are compared. The graph shows that WSN coverage optimization methods based on the particle swarm optimization (PSO), cosine function interactive dynamic adjustment particle swarm optimization (CPSO), nonlinear interactive adjustment particle swarm optimization (FPSO), uniform search particle swarm optimization (UPSO), and linear interactive dynamic adjustment particle swarm optimization (XPSO) all achieve a certain improvement in sensor node coverage, but some coverage blind spots still exist, resulting in low coverage. However, coverage optimization using the VGPSO algorithm guided by the Halton sequence and Voronoi diagram achieves a more even distribution of sensor nodes, reduces overlap, and achieves higher coverage than the previous methods.

[0166] In the process of optimizing wireless sensor network (WSN) coverage using VGPSO, the algorithm demonstrated remarkable convergence speed in the first 50 iterations, significantly outperforming other compared algorithms. After the 200th iteration, the convergence rate gradually leveled off. Ultimately, the VGPSO algorithm achieved an optimal coverage rate of 98.137% and an average coverage rate of 96.818%. Compared to the other five WSN coverage optimization algorithms, the VGPSO algorithm achieved higher coverage. Under the aforementioned experimental conditions, 20 cycles of testing were conducted on the six algorithms. The experimental results are shown in Table 1.

[0167] Table 1

[0168]

[0169] The above are merely exemplary embodiments of the present disclosure and are not intended to limit the scope of the present disclosure. In other words, any equivalent variations and modifications made in accordance with the teachings of the present disclosure are still within the scope of the present disclosure. Those skilled in the art will readily conceive of other embodiments of the present disclosure after considering the disclosure and the practical implications thereof.

[0170] This application is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary techniques in the art not described herein. The description and examples are to be considered as exemplary only, and the scope and spirit of the present disclosure are to be defined by the claims.

Claims

1. A wireless sensor network node configuration optimization method based on VGPSO algorithm, characterized in that: The method comprises the following steps: S1: Define wireless sensor network node configuration optimization as a problem ; S2: Improve the particle swarm algorithm through random Halton sequence, multi-factor cosine dynamic parameter adjustment strategy, multi-population multi-strategy evolution method and Voronoi diagram guidance; S3: Improve the particle swarm algorithm for problem Solve and output the global optimal position of each sensor node in the wireless sensor network; Step S3 includes: S31: Setup Issues The basic parameters and parameters of improved particle swarm algorithm; S32: Determine the optimal number of sensor nodes in the wireless sensor network; S33: Initialize the population using random Halton sequence; S34: Based on the cosine function, a control factor is introduced to design a multi-factor cosine dynamic parameter adjustment strategy; through the multi-factor cosine dynamic parameter adjustment strategy, the inertia weight and the global and individual learning factors are dynamically adjusted; S35: Using a multi-population and multi-strategy evolutionary approach, the initial population is divided into three sub-populations according to the preset ratio based on the fitness function values ​​of the individuals in the population; Adopting a multi-population and multi-strategy evolutionary approach, designing a speed update strategy for each sub-population and updating the sub-population; S36: Perform reflection out-of-bounds processing and position update on the updated sub-population; S37: Calculate the boot probability, Centroid and guide step length; S38: Update the position of the sensor node and the current individual fitness value through the guidance probability and guidance step size; compare the current individual fitness value of the population with the optimal fitness value of the historical evolution generation, update the optimal solution of the position of the sensor node until the maximum evolution generation is reached, and output the global optimal position of each sensor node.

2. A wireless sensor network node configuration optimization method based on the VGPSO algorithm according to claim 1, characterized in that: Step S31 includes: Defining the Problem The fitness function ,question The decision vector , Dimension ,as well as The length of the upper bound vector of the search region in the problem and the lower boundary vector , and the width upper boundary vector and the lower boundary vector ; Set the sensor's sensing radius to , number of nodes , the discrete granularity is 1; Determine the relevant parameters of the particle swarm algorithm, including: evolutionary generations Maximum speed , minimum speed , initialize the population size , control factor ; The problem M The search area in is discretized into pixels, if the wireless sensor network covers pixels, the coverage for .

3. A wireless sensor network node configuration optimization method based on the VGPSO algorithm as claimed in claim 2, characterized in that: Step S32 includes: set up and They represent the locations of three sensor nodes in the wireless sensor network; and The triangle formed by three points in a two-dimensional plane is an equilateral triangle; let A represent the center point of the equilateral triangle; Line segment is the radius , and As the center of the circle, construct three circles, point A is the intersection of the three circles, point B is the center of the circle and circle The intersection of Connecting line segments and line segments , the intersection of the two line segments is point C; According to the properties of the circle , calculate the optimal number of sensor nodes , the formula is as follows: in Represents a line segment length; Represents a line segment and line segments Angle; Represents a line segment length.

4. A wireless sensor network node configuration optimization method based on the VGPSO algorithm as claimed in claim 3, characterized in that: Step S33 includes: exist Within the scope, use Sequence initialization random generation Individuals, that is The solution vector of each individual is , where the odd-numbered bits range from ], the range of even-numbered digits is ; individual Together they constitute the parent population, and the formula for initializing the parent population is as follows: is The range of sequence generation values ​​is A randomly distributed variable, The formula for the process is as follows: in and Represent the upper and lower boundaries of odd-numbered digits, respectively. and Respectively represent the upper and lower boundaries of even digits; Indicates the first Individual odd-dimensional variables, Indicates the first Even-dimensional variables of individuals, and constitutes an individual ; To generate the The prime number of the cardinality of the dimension, Represents an integer In the cardinality The next number of digits; is the cardinality The next No. Dimensional Sequence value, is an integer In the cardinality The index of the highest digit under ; is an integer In the cardinality The next Number of digits.

5. A wireless sensor network node configuration optimization method based on VGPSO algorithm as claimed in claim 4, characterized in that: Step S34 includes: Introducing control factors , controlled by the factor and the current generation of the particle swarm algorithm calculate ; Use the cosine function to adjust the inertia weight , the formula is as follows: Using inertia weight and , calculate the global and individual learning factors and , the formula is as follows: in ; Indicates an intermediate parameter.

6. A wireless sensor network node configuration optimization method based on the VGPSO algorithm as claimed in claim 5, characterized in that: Step S35 includes: According to the fitness function value of the individuals in the population, the initial population is divided into three sub-populations according to the preset ratio , according to the fitness level, they are: development sub-population, moderate sub-population and exploration sub-population; Speed ​​update strategies include: elite learning strategy, coordinated learning strategy and generalized learning strategy; Development subpopulation Using elite learning strategy, moderate subpopulation Use coordinated learning strategies to explore subpopulations Use generalized learning strategies; The speed update formula of the elite learning strategy is as follows: The speed update formula of the coordinated learning strategy is as follows: The speed update formula of the generalized learning strategy is as follows: in is a random integer, For the The speed of the next generation of individuals, For the current The speed of an individual, and is an individual randomly selected from the parent population, It is one of the individuals in the population whose fitness function value ranks in the top preset percentage from small to large; is the current individual historical optimal solution; is a random number in the range of [-0.5,0.5], is the value range of each variable; and The random value range is [0,0.5].

7. A wireless sensor network node configuration optimization method based on the VGPSO algorithm as claimed in claim 6, characterized in that: Step S36 includes: If the speed If the speed exceeds the specified range, the reflection method with a certain interval is used to reflect the variable value back to the legal range. The formula of the reflection process is as follows: in The function is used to find the remainder of non-integer numbers. Indicates the minimum speed; Indicates the maximum speed; Update the position of the sensor node: according to the speed Calculate the position of the next generation of population individuals , for position Perform reflection out-of-bounds processing.

8. A wireless sensor network node configuration optimization method based on VGPSO algorithm as claimed in claim 7, characterized in that: Step S37 includes: The formula for calculating the bootstrap probability is as follows: in , used to control the steepness of the logistic function, represents the threshold that controls the maximum increase in the bootstrap probability, For continuous algebra without improvement; is the minimum bootstrap probability, is the maximum bootstrap probability; Indicates the Bootstrap probability of generation; According to the population position of the current generation, split Axis direction coordinate list and Y axis direction coordinate list, get the circle center list ; Calculate based on the circle center list centroid , the formula is as follows: in is the center of the circle list The X-axis coordinate of the coordinate, is the center of the circle list The Y-axis coordinate of the coordinate; Indicates the current The number of vertices of the polygon; Traverse the list of circle centers obtained by the population layout position, In the list, find the circle with the current center The distance is minimum and smaller than the perception radius The guiding centroid ); The formula for calculating the bootstrap step size is as follows: in is the current center and centroid distance; The center of the circle and The angle difference of the centroid, and are the guiding step length and guiding angle, is the guide vector; is a randomization function, cos and sin are trigonometric functions; is the function for calculating the angle.

9. A wireless sensor network node configuration optimization method based on the VGPSO algorithm as claimed in claim 8, characterized in that: Step S38 includes: The formula for updating the position of the sensor node is as follows: in, is the bootstrap probability; is the randomization function.

Images