WSN target coverage method and device based on chaos adaptive bat algorithm

Abstract

The invention discloses a wireless sensor network (WSN) target coverage method and device based on a chaos adaptive bat algorithm, and belongs to the technical field of wireless sensor network coverage. Aiming at the problems of limited communication range, poor adaptive capability and the like of the existing WSN, the method comprises the following steps of: initializing parameters, constructing a monitoring relation matrix to determine a fitness value, designing an adaptive inertia weight strategy, updating various parameters of a bat individual, and carrying out operations such as local search, adaptive adjustment of loudness and pulse emissivity and the like. And judging a termination condition to obtain an optimal solution in combination with an elitist strategy and chaos updating. The device comprises a network model construction module and the like and is used for executing the method. According to the method, the adaptive bat algorithm is adopted to establish the target coverage model, the elite strategy and chaos updating are combined, the algorithm is prevented from falling into local optimum, the stability is improved, the sensor node coverage area is effectively increased, the convergence speed is higher, the optimization capacity is stable, and the method is suitable for a WSN target coverage scene.

Description

Technical Field

[0001] The present invention relates to the field of Wireless Sensor Network (WSN), and in particular to a WSN target coverage method based on a chaotic adaptive bat algorithm, which is used for optimizing the coverage effect of sensor nodes in the WSN on targets. Background Art

[0002] Wireless sensor networks (WSNs) have widespread applications in numerous fields, including military surveillance, environmental monitoring, and industrial control. Ensuring effective coverage of target areas or points by sensor nodes is a key issue in these scenarios. Traditional sensor node deployment methods can suffer from coverage blind spots and uneven node energy consumption. While some existing algorithms can address some of these coverage issues, there is still room for improvement in complex environments and large-scale networks. The bat algorithm, a heuristic algorithm based on bat echolocation, offers certain advantages in optimizing targets. However, when applied to WSN target coverage, traditional bat algorithms suffer from slow convergence and a tendency to fall into local optima. Therefore, improvements are needed to meet the specific requirements of WSN target coverage. In many practical applications, the primary purpose of WSNs is to monitor specific targets or areas. Since the sensing range of sensor nodes is limited and constrained by various factors, such as node energy consumption, cost, and environmental interference, effectively deploying and scheduling sensor nodes to ensure adequate target coverage becomes a key issue.

[0003] However, existing technologies have the following problems: WSNs have limited communication range, lack of adaptability, susceptibility to interference, and difficulty responding to environmental changes and node failures. While bat-based algorithms for optimizing WSN node layout have potential, they are complex to set up, and network dynamics can affect their adaptability.

[0004] Therefore, a WSN target coverage method and device with comprehensive coverage, no coverage redundancy and strong algorithm adaptability is needed. Summary of the Invention

[0005] In view of this, the present invention provides a technical method to optimize the problem of incomplete coverage of WSN target coverage methods in the prior art, and utilizes a WSN target coverage method and device based on a chaotic adaptive bat algorithm to optimize the defects of the above-mentioned existing problems.

[0006] The specific technical solution of the present invention is as follows: a WSN target coverage method based on a chaotic adaptive bat algorithm, the main steps of which are as follows:

[0007] S1. Initialize parameters, set the population size (ie, the number of bats) to n, and the maximum number of iterations to Tmax , the search space dimension d, the search space area B, W is the adaptive inertia weight coefficient, the initial values ​​of loudness A and pulse emission rate r, determine the minimum frequency value f of the pulse frequency f emitted by the bat min and the maximum frequency f max , randomly initialize the position X of the bat population in the search space i (i=1,2,…,n), and speed V i (i=1,2,…,n);

[0008] S2. Taking the coverage rate C of sensor nodes as the optimization target, obtain the monitoring relationship matrix M containing the bat positions to represent the bat fitness;

[0009] The monitoring relationship matrix M is expressed as:

[0010] Among them, E is the monitored target, S is the sensor node, m e,s Indicates that the e-th monitored target is monitored by the s-th sensor node, m e,s = 0 means that the e-th monitored target is outside the coverage of the s-th sensor node, m e,s =1 means that the e-th monitored target is within the coverage of the s-th sensor node;

[0011] Due to the constraints of monitoring capabilities, sensor nodes can only select a limited number of monitored targets within the coverage area for monitoring. The monitoring relationship between sensor nodes and targets in wireless sensor networks can be expressed as:

[0012] In the monitoring relationship matrix Z, z e,s =1 means that the e-th monitored target is monitored by the s-th sensor node, and z e,s = 0 means that the e-th monitored target is within the coverage of the s-th sensor node but is not monitored, or the e-th monitored target is outside the coverage of the s-th sensor node. When each sensor node can monitor at most F targets within its coverage, the constraint condition is expressed as:

[0013] When the optimization goal is to successfully monitor the maximum number of monitored targets, that is, the coverage of sensor nodes is maximized, Fitness(X i ) function to obtain the target coverage model W e :

[0014] Among them, W eThe calculation relies on the monitoring relationship matrix Z, when a target e is detected by at least X sensor nodes, W e = 1, otherwise W e = 0. The specific calculation is as follows:

[0015] The constraint condition of the target coverage model W e is expressed as:

[0016] Wherein, W e = 0 indicates that the monitored target is outside the coverage range of the sensor node, W e = 1 indicates that the monitored target is within the coverage range of the sensor node, and X is the minimum value of the number of sensor nodes monitoring a single monitored target;

[0017] S3. Calculate the fitness value of the bat individual For each position X i of the bat individual, according to the specific target function to be optimized, the fitness value Fitness(X i ) is calculated:

[0018] Wherein W e (X i ) is the W e value determined based on the position X i of the bat individual, and the formula represents the sum of the number of successfully monitored targets under the monitoring relationship determined by the position of the bat individual. Therefore, the larger the fitness value, the more targets can be monitored by the sensor node deployment scheme corresponding to the position of the bat individual, and the better the scheme.

[0019] S4. Design an adaptive inertia weight strategy, and the adaptive inertia weight coefficient formula is expressed as:

[0020] Wherein, is the adaptive inertia weight coefficient of the i-th bat in the T-th iteration, is the maximum fitness value in all bat individuals in the T-th iteration, is the fitness value of the current i-th bat in the T-th iteration;

[0021] S5. Update the frequency of the bat individual;

[0022] The frequency of the bat individual i in the T-th iteration is updated according to the formula: ​

[0023] Among them, β is a random number uniformly distributed between [0,1], d ij is the distance from bat individual i to target j, is the average distance;

[0024] S6. Update the speed and position of the individual bat;

[0025] The speed of bat individual i at the Tth iteration The update formula is:

[0026] in, is the velocity of the previous iteration, is the position of the previous iteration, is the global optimal position in the previous iteration, and the elite bat set is E={e1,e 2, …,e k}, λ is the elite guidance factor, which ranges from [0, 1];

[0027] S7. Local search (random walk), update its position as follows: X new =X old +εA T

[0028] Among them, X new is the updated position, X old is the current position, ε is a random number uniformly distributed in the interval [-1,1], A T is the loudness at the current iteration;

[0029] S8. Update of loudness and pulse firing rate: If the fitness value of bat individual i at the Tth iteration is Than the fitness value of the previous iteration Improved, that is but:

[0030] Among them, α is the loudness attenuation coefficient, γ is the pulse emission rate increase coefficient, α, γ∈(0,1). On the contrary, if but:

[0031] S9. Fitness Assessment and Elite / Chaos Update: Calculate the fitness value of each bat individual after the update, update the elite bat individual set based on the fitness value, add newly generated bat individuals with higher fitness values ​​to the elite set, and remove elite individuals with lower fitness values;

[0032] Set the bat population density threshold δ, when λ 2 ≤δ, the results show that the algorithm is trapped in the local optimum and then changes the optimization strategy of the next generation iteration. All bat individuals no longer search randomly with the optimal position as the center, but use the Logistic chaos strategy to update the individual position so that it can jump out of the optimal value in the local range;

[0033] The individual position update formula after the normalization of the positions of k bats is as follows:

[0034] The updated position formula of the k+1th bat is as follows: x k+1 =μx k (1-x k ),0≤x k ≤1

[0035] Among them, μ is the chaos operator, μ=2, and the individual adaptability value of the bat is Fitness(X i ), λ is the elite guidance factor, in this formula, x k Obtained by normalizing the fitness value of the current individual bat. The choice of μ = 2 has been verified by research and practice. Under this value, the chaotic mapping can produce a relatively ideal chaotic sequence, making the position update of the individual bat in the search space more random and ergodic, effectively guiding the algorithm to escape the local optimum. Through this chaotic update method, the individual bat is no longer limited to a random search centered on the optimal position, but instead utilizes the characteristics of the chaotic sequence to explore in a wider space, increasing the possibility of finding the global optimal solution.

[0036] S10. Determine the termination condition;

[0037] Determine whether the maximum number of iterations T has been reached max If it has been reached, the iteration is stopped and the currently found global optimal solution (i.e., the optimal individual bat position and its corresponding fitness value) is output; if the maximum number of iterations has not been reached, it returns to S5 and continues to the next round of iterations, continuously optimizing to find a better solution.

[0038] This application also proposes a WSN target coverage device based on a chaotic adaptive bat algorithm, which is used to execute the method described in the technical solution, including a network model construction module, a fitness function design module, a chaotic adaptive bat algorithm module and a node deployment adjustment module;

[0039] The network model construction module receives input parameters, including parameters related to the monitoring area, such as the area shape (rectangular, circular, etc.) and area size. It also obtains parameters of the sensor nodes, including the sensing radius, communication radius, initial energy, etc. At the same time, it also needs to obtain specific parameters required by the chaotic adaptive bat algorithm, such as the initial value of the chaotic operator and related parameters of the chaotic sequence.

[0040] The network topology is established by constructing an initial distribution model of sensor nodes and targets within the monitoring area based on input parameters. The characteristics of chaotic sequences, such as randomness and ergodicity, are utilized to optimize the initial layout of nodes and targets, avoiding the problems of node clustering or irrational distribution that may result from traditional random initialization. Potential connections between nodes and between nodes and targets are determined to form an initial network topology, providing the basic data structure for subsequent calculations of node coverage of targets and network communication status.

[0041] The target coverage calculation part defines a function to calculate the number or proportion of targets covered under each sensor node deployment scheme. By traversing the sensing range of the sensor nodes and the target position, it is determined whether the target is within the sensing range of at least one node;

[0042] The comprehensive fitness function is constructed by combining factors such as target coverage and connectivity according to certain weights to form a fitness function. The characteristics of the chaos algorithm are used to dynamically adjust the weights of factors such as target coverage and connectivity. As the algorithm iterates, the weights are adaptively adjusted according to the changes in the chaotic sequence, making the fitness function more suitable for the search needs at different stages.

[0043] Bat individual initialization involves randomly generating a group of bat individuals using a chaotic initialization method based on the number of sensor nodes and the monitoring area determined in the network model building module. Each bat individual represents a sensor node deployment scheme, and its location information corresponds to the coordinates of the node within the monitoring area. The initial bat individual positions and velocities are generated through chaotic mapping, fully leveraging the diversity of chaotic sequences, increasing the variability of the initial population, improving the algorithm's search capabilities in the initial stages, and avoiding premature entrapment in local optima.

[0044] The fitness evaluation uses the fitness function in the fitness function design module to calculate the fitness value of each bat individual, which is used to evaluate the pros and cons of each deployment plan;

[0045] The chaotic elite selection selects a certain proportion of elite bat individuals based on their fitness values. These elite individuals have better fitness values ​​and represent a better node deployment solution.

[0046] The speed and position update is based on the basic principles of the bat algorithm and uses the information of elite individuals to update the speed and position of other bat individuals. An adaptive mechanism is introduced to dynamically adjust the pulse emission rate, loudness and other parameters of the bat individual according to the number of iterations of the algorithm and the current search status to balance the global search and local search capabilities. During the update process, ensure that the node position is within a reasonable range within the monitoring area. At the same time, when the algorithm determines that it may fall into a local optimum, the role of chaotic update is strengthened and the amplitude or frequency of chaotic disturbance is increased. A synergistic mechanism between chaotic strategy and elite strategy is established. Elite individuals are not only used to guide the update of ordinary bat individuals, but also serve as a reference point for chaotic update, so that bat individuals are more inclined to move closer to elite individuals during chaotic update, thereby balancing global search and local development capabilities;

[0047] The termination condition judgment sets appropriate termination conditions, such as reaching the maximum number of iterations, fitness value reaching a predetermined threshold, etc. When the termination condition is met, the current optimal bat individual, that is, the optimal sensor node deployment plan, is output.

[0048] The receiving optimal solution receives the final optimal sensor node deployment solution from the elite adaptive bat algorithm module;

[0049] The deployment instruction generation generates a deployment instruction for each sensor node according to the node position information in the optimal solution;

[0050] The deployment execution can interface with the actual WSN system to send deployment instructions to the corresponding sensor node actuators or adjust node positions and parameters in a virtual simulation environment. Verification and feedback are provided for the deployed network, introducing chaos analysis. By observing changes in the chaotic sequence during the evaluation of network coverage after deployment, it is determined whether further adjustments to node deployment are needed. If the chaotic sequence indicates that the network has areas with poor local coverage, targeted fine-tuning of node positions can be performed to ensure that node deployment meets expectations and that efficient target coverage is achieved. Ensure that node deployment meets expectations and that efficient target coverage is achieved.

[0051] The beneficial effects of the present application are as follows: the present application adopts an adaptive bat algorithm to establish a target coverage model to search for a sensor node deployment scheme; an elite strategy is adopted to enhance the optimization ability of the algorithm for the optimal individual, expand the population search range, effectively avoid falling into a local optimum during late iteration operation, and improve the stability of the algorithm; in terms of coverage effect, the WSN target coverage method based on the chaos adaptive bat algorithm effectively improves the sensor node coverage area. Compared with the prior art, the WSN target coverage method based on the chaos adaptive bat algorithm (CABA) effectively improves the sensor node coverage area, and has faster operation convergence speed and stable optimization ability. BRIEF DESCRIPTION OF DRAWINGS

[0052] Figure 1 A flowchart of the WSN target coverage method based on the chaos adaptive bat algorithm;

[0053] Figure 2 A structural diagram of the WSN target coverage device based on the chaos adaptive bat algorithm;

[0054] Figure 3 A simulation comparison diagram of the coverage ranges of three WSN target coverage methods;

[0055] Figure 4 A summary drawing of the WSN target coverage method and device based on the chaos adaptive bat algorithm. DETAILED DESCRIPTION

[0056] In order to make the technical solutions and advantages in the embodiments of the present application clearer, the exemplary embodiments of the present application will be further described in detail below with reference to the drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than an exhaustive enumeration of all embodiments. It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict.

[0057] Embodiment 1: Reference Figure 1 The WSN target coverage method based on the chaos adaptive bat algorithm specifically includes the following steps:

[0058] S1. Initialize parameters, set the population size (i.e. the number of bats) to n = 40, the maximum number of iterations T = 100, the search space dimension d = 2, W is the adaptive inertia weight coefficient, the initial values of the loudness A and the pulse emission rate r, determine the minimum frequency value f and the maximum frequency value f of the pulse frequency f emitted by the bat, and randomly initialize the position X of the bat population in the search space. max min max i ​​​(i=1,2,…,n), and speed V i (i=1,2,…,n);

[0059] S2. Taking the coverage rate C of sensor nodes as the optimization target, obtain the monitoring relationship matrix M containing the bat positions to represent the bat fitness;

[0060] The monitoring relationship matrix M is expressed as:

[0061] Among them, E is the monitored target, S is the sensor node, m e,s Indicates that the e-th monitored target is monitored by the s-th sensor node, m e,s = 0 means that the e-th monitored target is outside the coverage of the s-th sensor node, m e,s =1 means that the e-th monitored target is within the coverage of the s-th sensor node;

[0062] Due to the constraints of monitoring capabilities, sensor nodes can only select a limited number of monitored targets within the coverage area for monitoring. The monitoring relationship between sensor nodes and targets in wireless sensor networks can be expressed as:

[0063] In the monitoring relationship matrix Z, z e,s =1 means that the e-th monitored target is monitored by the s-th sensor node, and z e,s = 0 means that the e-th monitored target is within the coverage of the s-th sensor node but is not monitored, or the e-th monitored target is outside the coverage of the s-th sensor node. When each sensor node can monitor at most F targets within its coverage, the constraint condition is expressed as:

[0064] When the optimization goal is to successfully monitor the maximum number of monitored targets, that is, the coverage of sensor nodes is maximized, Fitness(X i ) function to obtain the target coverage model W e :

[0065] Among them, W e The computational dependency and monitoring relationship matrix Z, when a target e is detected by at least X sensor nodes, W e =1, otherwise W e =0. The specific calculation is as follows:

[0066] Target coverage model W eThe constraints are expressed as:

[0067] S3. Calculate the fitness value of the bat individual for each bat individual's location X i , calculate its fitness value Fitness(X according to the specific objective function to be optimized i The fitness value reflects the quality of the bat's individual position. In the minimization problem, the smaller the fitness value, the better; in the maximization problem, the larger the fitness value, the better.

[0068] S4. Design an adaptive inertia weight strategy. The adaptive inertia weight coefficient formula is expressed as:

[0069] in, is the adaptive inertia weight coefficient of the i-th bat in the T-th iteration, is the maximum fitness value of all bat individuals in T iterations, is the fitness value of the current i-th bat in T iterations;

[0070] S5. Frequency of updating bat individuals;

[0071] The frequency of bat individual i at the Tth iteration The update formula is:

[0072] Among them, f min and f max are the minimum and maximum values ​​of the frequency, β is a random number uniformly distributed between [0,1], and d ij is the distance from bat individual i to target j, is the average distance. By simulating the bat's behavior of emitting pulses of different frequencies at different stages, the bat can dynamically adjust its search direction and speed based on information such as the distance of the target during the search process, making it easier to find a high-quality solution quickly.

[0073] S6. Update the speed and position of the individual bat;

[0074] The speed of bat individual i at the Tth iteration The update formula is:

[0075] in, is the velocity of the previous iteration, is the position of the previous iteration, is the global optimal position in the previous iteration, and the elite bat set is E={e1,e2, …,e k}, λ is the elite guidance factor, ranging from [0 to 1]. Based on the updated frequency, the speed and position of individual bats are updated according to this improved formula. During the update process, an elitist strategy is introduced, which preserves the best bats (those with higher fitness values) in each generation and allows these elite individuals to guide the updates of other individuals. For example, when ordinary bats update their positions, they can move closer to elite individuals with a certain probability, thereby accelerating the algorithm's convergence and preventing the algorithm from falling into a local optimum too early.

[0076] S7. Local search (random walk), update its position as follows: X new =X old +εA T

[0077] Among them, X new is the updated position, X old is the current position, ε is a random number uniformly distributed in the interval [-1,1], A T is the loudness at the current iteration. This process simulates the behavior of bats exploring a small area nearby after finding a local optimal area, which helps to further explore local optimal solutions.

[0078] S8. Update of loudness and pulse firing rate: If the fitness value of bat individual i at the Tth iteration is Than the fitness value of the previous iteration Improved, that is but:

[0079] Among them, α is the loudness attenuation coefficient, γ is the pulse emission rate increase coefficient, α, γ∈(0,1). On the contrary, if but:

[0080] Adaptively adjust the loudness and pulse emission rate of individual bats based on their fitness. Bats with high fitness values ​​have their loudness gradually decrease and their pulse emission rate gradually increase, which makes good solutions more influential in subsequent searches while reducing the interference of poor solutions on the search process.

[0081] S9. Fitness evaluation and elite and chaotic updates: Calculate the fitness value of each bat individual after the update, update the elite bat individual set based on the fitness value, add the newly generated bat individuals with higher fitness values ​​to the elite set, and remove the elite individuals with lower fitness values.

[0082] Set the bat population density threshold δ, when λ 2 ≤δ, the results show that the algorithm is trapped in the local optimum and then changes the optimization strategy for the next iteration. All bat individuals no longer randomly search around the optimal position, but use the Logistic Chaos strategy to update their individual positions, so that they can jump out of the optimal value in the local range.

[0083] The individual position update formula after the normalization of the positions of k bats is as follows:

[0084] The updated position formula of the k+1th bat is as follows: x k+1 =μx k (1-x k ),0≤x k ≤1

[0085] Among them, μ is the chaos operator, μ=2, and the individual adaptability value of the bat is Fitness(X i ), λ is the elite guidance factor, in this formula, x k The value is normalized by the fitness of the current individual bat. The choice of μ = 2 is validated by research and practice. At this value, the chaotic mapping produces a relatively ideal chaotic sequence, ensuring that the position updates of individual bats in the search space are both random and ergodic, effectively guiding the algorithm out of local optima. This chaotic update method eliminates the need for random searches centered on the optimal position. Instead, it leverages the properties of chaotic sequences to explore a wider range of space, increasing the likelihood of finding the global optimal solution.

[0086] S10. Determine the termination condition;

[0087] Determine whether the maximum number of iterations T has been reached max If it has been reached, the iteration is stopped and the currently found global optimal solution (i.e., the optimal individual bat position and its corresponding fitness value) is output; if the maximum number of iterations has not been reached, it returns to S5 and continues to the next round of iterations, continuously optimizing to find a better solution.

[0088] Specifically, in this implementation, the detection range of the WSN target coverage method based on the chaotic adaptive bat algorithm is set to 600m × 600m, the number of monitored targets is 200, and the locations of the monitored targets and sensor nodes are randomly distributed. Each monitored target requires at least three sensor nodes to monitor simultaneously, and each sensor node can simultaneously sense up to five targets within the coverage range. The number of sensors is 120, the sensor sensing radius is 60m, the bat population size is 40, and the number of iterations is 100.

[0089] refer to Figure 3 The WSN target coverage method based on the Chaos Adaptive Bat Algorithm was compared with the Genetic Algorithm and the Fruit Fly Algorithm. The population size of each algorithm was 40, the number of iterations was 100, and the detection range was 600m × 600m. The Fruit Fly Algorithm's search step was 10. The Genetic Algorithm used a roulette wheel selection method with a crossover probability of 0.8 and a mutation probability of 0.1. Simulation curves show that the WSN target coverage method based on the Chaos Adaptive Bat Algorithm reaches a plateau after the 60th iteration, while the Genetic Algorithm and the Fruit Fly Algorithm converge too quickly. Furthermore, the WSN target coverage method based on the Chaos Adaptive Bat Algorithm achieved a coverage rate of over 85%, while the Fruit Fly Algorithm's coverage rate was around 70%, and the Genetic Algorithm's coverage rate was around 60%. Compared to the Genetic Algorithm and the Fruit Fly Algorithm, the WSN target coverage method based on the Chaos Adaptive Bat Algorithm achieved a higher coverage rate.

[0090] Example 2: Reference Figure 2 Detailed description of this embodiment, a WSN target coverage device based on a chaotic adaptive bat algorithm, for executing the method described in Example 1, comprising a network model construction module, a fitness function design module, a chaotic adaptive bat algorithm module, and a node deployment adjustment module;

[0091] The network model construction module receives input parameters, including parameters related to the monitoring area, such as the area shape (rectangular, circular, etc.) and area size. It also obtains parameters of the sensor nodes, including the sensing radius, communication radius, initial energy, etc. At the same time, it also needs to obtain specific parameters required by the chaotic adaptive bat algorithm, such as the initial value of the chaotic operator and related parameters of the chaotic sequence.

[0092] The network topology is established by constructing an initial distribution model of sensor nodes and targets within the monitoring area based on input parameters. The characteristics of chaotic sequences, such as randomness and ergodicity, are utilized to optimize the initial layout of nodes and targets, avoiding the problems of node clustering or irrational distribution that may result from traditional random initialization. Potential connections between nodes and between nodes and targets are determined to form an initial network topology, providing the basic data structure for subsequent calculations of node coverage of targets and network communication status.

[0093] The target coverage calculation part defines a function to calculate the number or proportion of targets covered under each sensor node deployment scheme. By traversing the sensing range of the sensor nodes and the target position, it is determined whether the target is within the sensing range of at least one node;

[0094] The comprehensive fitness function is constructed by combining factors such as target coverage and connectivity according to certain weights to form a fitness function. The characteristics of the chaos algorithm are used to dynamically adjust the weights of factors such as target coverage and connectivity. As the algorithm iterates, the weights are adaptively adjusted according to the changes in the chaotic sequence, making the fitness function more suitable for the search needs at different stages.

[0095] Bat individual initialization involves randomly generating a group of bat individuals using a chaotic initialization method based on the number of sensor nodes and the monitoring area determined in the network model building module. Each bat individual represents a sensor node deployment scheme, and its location information corresponds to the coordinates of the node within the monitoring area. The initial bat individual positions and velocities are generated through chaotic mapping, fully leveraging the diversity of chaotic sequences, increasing the variability of the initial population, improving the algorithm's search capabilities in the initial stages, and avoiding premature entrapment in local optima.

[0096] The fitness evaluation uses the fitness function in the fitness function design module to calculate the fitness value of each bat individual, which is used to evaluate the pros and cons of each deployment plan;

[0097] The chaotic elite selection selects a certain proportion of elite bat individuals based on their fitness values. These elite individuals have better fitness values ​​and represent a better node deployment solution.

[0098] The speed and position update is based on the basic principles of the bat algorithm and uses the information of elite individuals to update the speed and position of other bat individuals. An adaptive mechanism is introduced to dynamically adjust the pulse emission rate, loudness and other parameters of the bat individual according to the number of iterations of the algorithm and the current search status to balance the global search and local search capabilities. During the update process, ensure that the node position is within a reasonable range within the monitoring area. At the same time, when the algorithm determines that it may fall into a local optimum, the role of chaotic update is strengthened and the amplitude or frequency of chaotic disturbance is increased. A synergistic mechanism between chaotic strategy and elite strategy is established. Elite individuals are not only used to guide the update of ordinary bat individuals, but also serve as a reference point for chaotic update, so that bat individuals are more inclined to move closer to elite individuals during chaotic update, thereby balancing global search and local development capabilities;

[0099] The termination condition judgment sets appropriate termination conditions, such as reaching the maximum number of iterations, fitness value reaching a predetermined threshold, etc. When the termination condition is met, the current optimal bat individual, that is, the optimal sensor node deployment plan, is output.

[0100] The receiving optimal solution receives the final optimal sensor node deployment solution from the elite adaptive bat algorithm module;

[0101] The deployment instruction generation generates a deployment instruction for each sensor node according to the node position information in the optimal solution;

[0102] The deployment execution can interface with the actual WSN system to send deployment instructions to the corresponding sensor node actuators or adjust node positions and parameters in a virtual simulation environment. Verification and feedback are provided for the deployed network, introducing chaos analysis. By observing changes in the chaotic sequence during the evaluation of network coverage after deployment, it is determined whether further adjustments to node deployment are needed. If the chaotic sequence indicates that the network has areas with poor local coverage, targeted fine-tuning of node positions can be performed to ensure that node deployment meets expectations and that efficient target coverage is achieved. Ensure that node deployment meets expectations and that efficient target coverage is achieved.

[0103] Although the present invention has been described with respect to a limited number of embodiments, it will be apparent to those skilled in the art, having benefit of the foregoing description, that other embodiments are contemplated within the scope of the invention thus described. Furthermore, it should be noted that the language used in this specification has been selected primarily for readability and didactic purposes, rather than for the purpose of explaining or limiting the subject matter of the present invention. Consequently, many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the appended claims. The disclosure of the present invention is intended to be illustrative rather than restrictive of the scope of the invention, which is defined by the appended claims.

Claims

1. A WSN target coverage method based on chaotic adaptive bat algorithm is characterized by: The following steps are involved: S1. Initialization: Set the population size, search space dimension, etc., set the initial values ​​of the adaptive inertia weight coefficient, loudness and pulse emission rate, and frequency range; randomly initialize the position and velocity of the bat population within the search space; randomly generate bat populations based on the parameter range of sensor nodes, use sensor node coverage as the optimization target, and obtain the monitoring relationship matrix; S2. Calculate the fitness value of individual bats: For each individual bat's location, calculate its fitness value based on the optimization objective function, design an adaptive inertia weight strategy, and obtain the adaptive inertia weight coefficient formula; S3. Update the frequency, speed, and position of individual bats; S4. Local search; S5. Update of loudness and pulse firing rate; S6. Fitness evaluation and elite and chaotic updates; S7. Determine the termination condition: Determine whether the maximum number of iterations has been reached. If so, stop the iteration and output the global optimal solution; otherwise, return to step S3 to continue the iteration.

2. The WSN target coverage method based on the chaotic adaptive bat algorithm according to claim 1 is characterized in that: The monitoring relationship matrix M is expressed as: Among them, E is the monitored target, S is the sensor node, m e,s Indicates that the e-th monitored target is monitored by the s-th sensor node, m e,s = 0 means that the e-th monitored target is outside the coverage of the s-th sensor node, m e,s =1 means that the e-th monitored target is within the coverage of the s-th sensor node; The constraints of the monitoring relationship matrix M are expressed as: Where F is the maximum number of targets that each sensor node can monitor within its coverage area.

3. The WSN target coverage method based on the chaotic adaptive bat algorithm according to claim 1 is characterized in that: The target coverage model W e Expressed as: The target coverage model W e The constraints are expressed as: Among them, W e =0 means the monitored target is outside the coverage of the sensor node, W e =1 means that the monitored target is within the coverage of the sensor node, and X is the minimum number of sensor nodes that monitor a single monitored target; e,s ∈{0,1}, represents the monitoring relationship between the e-th monitored target and the s-th sensor node, z e,s =1 means that the e-th monitored target is monitored by the s-th sensor node, and z e,s =0 means that the e-th monitored target is within the coverage of the s-th sensor node but is not monitored, or the e-th monitored target is outside the coverage of the s-th sensor node.

4. The WSN target coverage method based on the chaotic adaptive bat algorithm according to claim 1 is characterized in that: The adaptive inertia weight coefficient formula is expressed as: in, is the adaptive inertia weight coefficient of the i-th bat in the T-th iteration, is the maximum fitness value of all bat individuals in T iterations, is the fitness value of the current i-th bat in T iterations.

5. The WSN target coverage method based on the chaotic adaptive bat algorithm according to claim 1 is characterized in that: The frequency of the bat individual i at the Tth iteration The update formula is: Among them, f min and f max are the minimum and maximum values ​​of the frequency, β is a random number uniformly distributed between [0,1], and d ij is the distance from bat individual i to target j, is the average distance.

6. The WSN target coverage method based on the chaotic adaptive bat algorithm according to claim 1 is characterized in that: The speed of bat individual i at the Tth iteration The update formula is: in, is the velocity of the previous iteration, is the position of the previous iteration, is the global optimal position in the previous iteration, and the elite bat set is E={e1,e 2, …,e k }, λ is the elite guidance factor, which ranges from [0, 1].

7. The WSN target coverage method based on the chaotic adaptive bat algorithm according to claim 1 is characterized in that: In the local search operation, the position update formula is: X new =X old +εA T Among them, X new is the updated position, X old is the current position, ε is a random number uniformly distributed in the interval [-1,1], A T is the loudness at the current iteration.

8. The WSN target coverage method based on chaotic adaptive bat algorithm according to claim 1 is characterized in that: The loudness and pulse rate updates include: If the fitness value of bat individual i at the Tth iteration is Than the fitness value of the previous iteration Improved, that is but: On the contrary, if but: Among them, α is the loudness attenuation coefficient, γ is the pulse emission rate increase coefficient, α, γ∈(0,1).

9. The WSN target coverage method based on chaotic adaptive bat algorithm according to claim 1 is characterized in that: The fitness evaluation and elite and chaos updates include: Calculate the fitness value of each bat individual after the update, update the elite bat individual set according to the fitness value, add the newly generated bat individuals with higher fitness values ​​to the elite set, and remove the elite individuals with lower fitness values; set the bat population density threshold δ, when λ 2 ≤δ, the Logistic chaos strategy is used to update the individual position. The individual position update formula after the normalization of the k bat positions is as follows: The updated position formula of the k+1th bat is as follows: x k+1 =μx k (1-x k ),0≤x k ≤1 Where μ is the chaos operator, μ=2.

10. A WSN target coverage device based on chaotic adaptive bat algorithm, characterized in that: Used to execute the method according to any one of claims 1 to 8, comprising a network model construction module, a fitness function design module, a chaotic adaptive bat algorithm module and a node deployment adjustment module; The network model construction module is used to receive relevant parameters of the monitoring area (such as area shape and area size), parameters of sensor nodes (sensing radius, communication radius, initial energy, etc.), and specific parameters required by the chaotic adaptive bat algorithm (initial value of the chaotic operator, relevant parameters of the chaotic sequence, etc.), build an initial distribution model of sensor nodes and targets in the monitoring area, optimize the initial layout of nodes and targets using the chaotic sequence, determine the potential connection relationship between nodes and between nodes and targets, form an initial network topology, and calculate the target coverage rate; The fitness function design module is used to combine factors such as target coverage and connectivity according to certain weights to form a fitness function, and dynamically adjust the weights of factors such as target coverage and connectivity using the characteristics of the chaos algorithm; The chaotic adaptive bat algorithm module is used to randomly generate bat individuals using a chaotic initialization method based on the number of sensor nodes and monitoring area determined by the network model construction module, calculate the fitness value of each bat individual, select elite bat individuals, update the speed and position of other bat individuals based on the information of the elite individuals, introduce adaptive mechanisms (including adaptive inertia weight strategy, dynamic adjustment of pulse emission rate and loudness, etc.) to adjust parameters, determine termination conditions and output the optimal sensor node deployment plan; The node deployment adjustment module is used to receive the optimal sensor node deployment plan, generate deployment instructions for each sensor node, and send the deployment instructions to the corresponding sensor node actuator or adjust the node position and parameters in a virtual simulation environment. It verifies and provides feedback on the deployed network, introduces chaos analysis, and determines whether node deployment needs to be adjusted based on changes in the chaotic sequence.

Images