A Multi-Target Wireless Sensor Network Coverage Method Based on Hybrid Gray Wolf-Adaptive Butterfly Algorithm
By optimizing wireless sensor networks using a hybrid gray wolf-adaptive butterfly algorithm, the problems of coverage gaps and high energy consumption are solved, resulting in more efficient network coverage and energy management, and extending network lifespan.
Patent Information
- Application Number
- CN202310473697.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-04-27
- Publication Date
- 2025-11-14
- Estimated Expiration
- 2043-04-27
AI Technical Summary
Wireless sensor networks suffer from coverage gaps or redundancy during deployment, and have high energy consumption, making it difficult to extend network lifespan in harsh environments.
A hybrid gray wolf-adaptive butterfly algorithm is adopted, which combines node redundancy rate and energy consumption as optimization objectives, improves the adaptive switching probability and feedback mechanism, optimizes the sensor network coverage model, and performs global and local searches through the hybrid gray wolf-adaptive butterfly algorithm to prevent local convergence.
It improves the coverage and algorithm convergence speed of wireless sensor networks, reduces the number of nodes used and energy consumption, and extends network lifespan.
Smart Images
Figure CN116634449B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to a multi-target wireless sensor network coverage method based on the hybrid gray wolf-adaptive butterfly algorithm, belonging to the field of wireless sensor network coverage optimization technology. Background Technology
[0002] Wireless sensor networks (WSNs) have been widely adopted due to their low cost and ability to be deployed in harsh environments. However, because their deployment areas are often inaccessible and lack infrastructure, they are typically deployed randomly using aircraft or other means, resulting in high randomness and potential coverage gaps or redundancy. The coverage rate of a WSN determines its future performance; higher coverage leads to better sensing capabilities. Existing WSN coverage technologies typically focus on coverage rate as a single objective. Reducing the number of nodes used during the coverage process can help lower costs, and minimizing their travel distance can save energy and extend network lifespan. Therefore, multi-objective coverage optimization for WSNs is of great significance.
[0003] The technology in this application originates from the Yunnan Provincial Science and Technology Plan Major Science and Technology Special Project (202202AD080008); the Yunnan Provincial Basic Research Plan Key Project (202101AS070016); the Yunnan Provincial Key Laboratory of Computer Technology Application Open Fund (CB22144S073A); the 2020 Yunnan Provincial Postgraduate Excellent Course "Algorithm Analysis and Design" Construction Project (109920210048); the "Xingdian Elite Talent Support Program" Industrial Innovation Talent Project (Yunfa Gairenshi
[2019] No. 1096); and the Yunnan Provincial Technological Innovation Talent Project (2019HB113). Summary of the Invention
[0004] The technical problem this invention aims to solve is to provide a multi-target wireless sensor network coverage algorithm based on the hybrid gray wolf-adaptive butterfly algorithm. This algorithm can optimize the quality of wireless sensor networks, improve sensing capabilities, reduce the number of nodes used, and save costs, thereby solving the aforementioned technical problems. Furthermore, due to the harsh conditions and lack of infrastructure in network deployment areas, energy replenishment is difficult. Shortening the travel distance during secondary deployment can reduce energy consumption and extend network lifespan.
[0005] The technical solution of this invention is: a multi-target wireless sensor network coverage method based on a hybrid gray wolf-adaptive butterfly algorithm. First, in the traditional Boolean model of coverage, node redundancy and energy consumption are added as optimization objectives to optimize the coverage model. Second, the gray wolf algorithm is introduced into the global search of the traditional butterfly algorithm to help with optimization, integrating intra-population optimization on the basis of inter-population optimization to improve the convergence speed and optimization ability of the algorithm. Then, the adaptive switching probability is improved in the butterfly algorithm to determine the optimization method according to the current iteration, making reasonable use of computing resources and accelerating the search speed. Finally, an improved feedback mechanism is introduced into the improved butterfly algorithm, introducing random butterflies for walking to prevent getting trapped in local optima.
[0006] The specific steps are as follows:
[0007] Step 1: Establish a model to optimize multi-target coverage of wireless sensor networks from three aspects: minimum coverage, minimum node redundancy, and minimum energy consumption. Specifically:
[0008] Step 1.1: Establish the coverage function.
[0009] Assume the WSN deployment area is a rectangle with length L and width W, and n identical sensing sensors with a radius of r are placed within the area. s The sensors, where the i-th sensor node is denoted as s. i The area is divided into S monitoring points, S = L × W, and each monitoring point is denoted as m1, ..., m2. i ···m S Let m be the monitoring point. i Coordinates are (x node_i ,y node_i Sensor coordinates s i For (x) i ,y i If the distance between the sensor node and the point to be monitored is:
[0010]
[0011] Based on the Boolean model and the Euclidean distance between two points, the probability that a point to be monitored is covered by a sensor node is defined as follows:
[0012]
[0013] The joint sensing probability that the point to be monitored is covered by all sensor nodes is:
[0014]
[0015] The overall coverage of the wireless sensor network should be:
[0016]
[0017] Step 1.2: Estimate the maximum number of nodes in the network.
[0018] The estimation of the maximum number of network nodes is primarily for the purpose of rationally allocating the number of nodes in experimental design and practical applications, avoiding resource waste. During deployment, when a node's coverage area utilization rate is greater than 50%, the node should be deployed. To achieve near-complete coverage, the estimated maximum number of nodes to deploy within the area should be:
[0019]
[0020]
[0021] N = N L *N W (7)
[0022] Step 1.3: Establish the node redundancy rate function.
[0023] The node redundancy rate is defined as the ratio of the actual coverage area of a sensor to its capable coverage area. Let there be n identical sensors in the region with a sensing radius of r. s The node redundancy rate of the sensor is calculated as follows:
[0024]
[0025] Step 1.4: Establish the energy consumption function and build a multi-objective coverage model.
[0026] During the secondary deployment, nodes will move; the greater the movement distance, the higher the energy consumption. Summing and normalizing the movement distances of the sensor nodes yields the following energy consumption function:
[0027]
[0028] The multi-target coverage model is as follows:
[0029] f t =ω1(1-C r )+ω2p r +ω3p e (10).
[0030] In language identification, extracting key language-related features and removing minor features can effectively improve the accuracy of language identification. This method can improve the accuracy of language identification in broadcast audio and has a relatively fast identification speed.
[0031] Step 2: Initialize the population matrix, calculate the fitness of each population according to the model function, and then calculate its aroma intensity, specifically:
[0032] Step 2.1: Initialize the population matrix with dimensions pop × 2N, where pop is the population size and N is the number of nodes. The initialization formula is as follows:
[0033] X pop·2N =lb+rand(pop,2N)·(ub-lb) (11)
[0034] Among them, X pop·2N Let pop be the initial position matrix of all butterfly populations, 2N be the size of the butterfly population, and rand(pop,2N) be a random matrix of pop rows and 2N columns for each element value in the range [0,1]. ub and lb are the upper and lower bounds. Each row vector in the matrix represents the position of a group of sensor nodes.
[0035] Step 2.2: Calculate the fragrance intensity of each butterfly population using the following formula.
[0036] f = c·I α (12)
[0037] Where f is the intensity of the fragrance that other butterflies can perceive, c is the sensory morphology coefficient, I is the stimulus intensity, i.e. the function fitness value, which is obtained by substituting the node position of each butterfly population into the coverage model for calculation, and α reflects the degree of absorption, with a value between [0,1].
[0038] The sensory morphology coefficient c ranges from [0,∞), and is calculated as follows:
[0039]
[0040] Among them, T max This represents the maximum number of iterations.
[0041] Step 3: Perform a random walk on the randomized individual using an improved feedback mechanism. If the fitness of the randomized individual is better, move it in the direction jointly guided by the randomized individual and the best individual; otherwise, move it only in the direction guided by the best individual. Specifically:
[0042] In the butterfly algorithm's search process, each butterfly population can sense the scents emitted by other populations and will gravitate towards the better populations, performing a global search. However, due to the limited range of the butterfly's perception, when it cannot detect a better population, it will perform a random walk, performing a local search. In the traditional butterfly algorithm, the on / off probability is 0.8, and the algorithm randomly selects a value between 0 and 1. If the random number is less than the on / off probability, the algorithm performs a global search; otherwise, it performs a local search. Therefore, the algorithm is mostly guided by the optimal population and may get stuck in local convergence.
[0043] The feedback mechanism adjusts the guidance direction by introducing random butterflies, enabling it to perceive more information, enhancing its global awareness, and preventing the algorithm from getting stuck in local convergence. When the fitness of the random butterfly is better than that of the current butterfly, it moves under the joint guidance of both; otherwise, it moves towards the optimal butterfly population. The feedback mechanism was further improved by adjusting its scaling factor. The improved feedback mechanism adjustment method is as follows:
[0044]
[0045]
[0046]
[0047] In the formula, β is the sharing coefficient and γ is the feedback factor.
[0048] Step 4: Calculate the adaptive switching probability. Based on the magnitude of the random value and the adaptive switching probability, determine the optimization method. Use the butterfly algorithm of hybrid gray wolves to optimize the population position and handle out-of-bounds population behavior. Specifically:
[0049] Step 4.1: Calculate the adaptive switching probability.
[0050] In the traditional butterfly algorithm, experiments have shown that it is effective for most applications when the switching probability is 0.8, but it lacks specificity for certain applications during the iteration process. To make the iteration method more reasonable, this invention adaptively improves the switching probability, allowing it to determine the search method based on the current iteration status. The improved adaptive switching probability is:
[0051]
[0052] in, Let be the switching probability of the i-th population at the t-th iteration. f i t , These are the optimal fitness value, the i-th population fitness value, and the worst fitness value at the t-th iteration, respectively.
[0053] Step 4.2: Iteratively optimize the population matrix using the butterfly algorithm of the hybrid gray wolf algorithm.
[0054] When the random number is less than the adaptive switching probability, a global search is performed; otherwise, a local search is performed. Since the algorithm spends most of its time performing a global search, improving this method can significantly enhance the algorithm's optimization ability and convergence speed. The improved search method is as follows:
[0055]
[0056]
[0057]
[0058]
[0059] in, Let g represent the spatial positions of the i-th, j-th, and k-th butterflies in the t-th iteration, where j and k are randomly selected. * f represents the current optimal butterfly population position. i Let X be the fitness value of the i-th butterfly, r be a random number between [0,1], g1, g2, and g3 be the best, second-best, and third-best individuals in the current population, respectively. k (i,j) represents the coordinates of the j-th individual in the i-th population after being guided by the best, second-best, and third-best individuals, respectively, where i = 1, 2, ..., N, N is the population size, and j = 1, 2, ..., d, d is the spatial dimension.
[0060] If the improved algorithm performs a global search, it first executes formulas (18) to (20), and then executes the first formula of formula (21) to perform a global search. If it performs a local search, it executes the second formula of formula (21) to perform a local search.
[0061] Step 4.3: Handle the new population for crossing boundaries.
[0062] The conventional approach to handling individuals that have crossed the boundary is to reassign them to the boundary. However, considering that this might reduce population diversity, a new method is used: some of the individuals that have crossed the boundary are reassigned to the boundary, while the rest are randomly reassigned. Specifically:
[0063]
[0064] Where x1 is the new position of x after processing, and r, r1, and r2 are random numbers between [0,1].
[0065] Step 5: Update the optimal population position and parameter c. If the fitness value reaches the optimal or the maximum number of iterations is reached, the optimal population position is the optimized sensor node position. Otherwise, repeat Step 3 and Step 4.
[0066] Specifically:
[0067] Step 5.1: Update the optimal population position and parameter c.
[0068] If the updated butterfly population has a lower fitness when substituted into the coverage model, then the population is replaced; otherwise, it remains unchanged. After all populations have been replaced, the optimal population is selected for updating. In addition, parameter c also needs to be updated, as shown in formula (13).
[0069] Step 5.2: Determine if the condition for exiting the loop has been met.
[0070] If the fitness value reaches the optimal value or the maximum number of iterations is reached, then the optimal population position is the optimized sensor node position; otherwise, repeat Step 3 and Step 4.
[0071] The beneficial effects of this invention are as follows: In the field of wireless sensor network coverage, it firstly improves coverage and convergence speed compared to existing improved algorithms. Furthermore, it takes into account the number of nodes used, node redundancy, and energy consumption during secondary deployment, comprehensively optimizing multi-target coverage. This reduces network costs while improving coverage in the traditional way, extending network lifespan for long-term planning. Attached Figure Description
[0072] Figure 1 This is a flowchart illustrating the implementation of the present invention;
[0073] Figure 2 This is a state diagram when the nodes of this invention are fully covered;
[0074] Figure 3 This is a parameter diagram of the experimental scenario of this invention;
[0075] Figure 4 This is the initial coverage map of 40 nodes within a 100*100 area according to the present invention;
[0076] Figure 5 This is a line graph comparing the present invention with other algorithms across 40 nodes within a 100*100 area;
[0077] Figure 6 This is the final deployment coverage map of the present invention with 40 nodes in a 100*100 area and compared with other comparison algorithms;
[0078] Figure 7 These are graphs showing the results of different algorithms in four different scenarios. Detailed Implementation
[0079] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.
[0080] Example 1: As Figure 1 As shown, a multi-target wireless sensor network coverage algorithm based on the hybrid gray wolf-adaptive butterfly algorithm is presented:
[0081] Step 1: Build a model to optimize the multi-target coverage of wireless sensor networks from three aspects: minimum coverage, minimum node redundancy, and minimum energy consumption.
[0082] First, establish the coverage function. The deployment area of the WSN is a rectangle with length L and width W, and n identical sensing units with a radius of r are placed within the area. s The sensors, where the i-th sensor node is denoted as s. i Let m be the monitoring point. i Coordinates are (x node_i ,y node_i Sensor coordinates s i For (x) i ,y i If the distance between the sensor node and the point to be monitored is:
[0083]
[0084] The probability that the point to be monitored is covered by sensor nodes is:
[0085]
[0086] The joint sensing probability that the monitoring point is covered by all sensor nodes is:
[0087]
[0088] Continuing this deduction, the overall coverage of the wireless sensor network should be:
[0089]
[0090] Next, a node redundancy rate function is established. The node redundancy rate is defined as the ratio of the actual coverage area of a sensor to its capable coverage area. Let there be n identical sensing sensors within the region, with a sensing radius of r. s The specific method for calculating the node redundancy rate of the sensor is as follows:
[0091]
[0092] Finally, an energy consumption function and a multi-target coverage model are established. During the secondary deployment process, nodes will move; the farther the movement distance, the higher the energy consumption. The energy consumption function is obtained by summing and normalizing the movement distances of the sensor nodes:
[0093]
[0094] In summary, the multi-target coverage model is as follows:
[0095] f t =ω1(1-C r )+ω2pr +ω3p e (10)
[0096] Next, according to Figure 2 Estimate the maximum number of nodes in the network. During deployment, deploy a node when its coverage area utilization is greater than 50%. In summary, to achieve near-complete coverage, the estimated maximum number of nodes to deploy in the area should be:
[0097]
[0098]
[0099] N = N L *N W (7)
[0100] Step 2: Initialize the population matrix. The matrix dimension is pop × 2N, where pop is the population size and N is the number of nodes. The specific initialization formula is as follows:
[0101] X pop·2N =lb+rand(pop,2N)·(ub-lb) (11)
[0102] Among them, X pop·2N Let `pop` be the initial position matrix of all butterfly populations, `pop` be the butterfly population size, `2N` be the spatial dimension, and `rand(pop,2N)` be a random matrix of rows `pop` and columns `2N` where each element has a value in the range [0,1]. `ub` and `lb` are the upper and lower bounds. Each row vector in the matrix represents the position of a group of sensor nodes. Figure 4 for Figure 3 The initial deployment coverage map of 40 nodes in a 100*100 area is based on formula (11).
[0103] Next, the fragrance intensity of each butterfly population is calculated using the following formula:
[0104] f = c·I α (12)
[0105] Where f is the intensity of the fragrance that other butterflies can perceive, c is the sensory morphology coefficient, I is the stimulus intensity, i.e. the function fitness value, which is obtained by substituting the node position of each butterfly population into the coverage model for calculation, and α reflects the degree of absorption, with a value between [0,1].
[0106] The sensory morphology coefficient c ranges from [0,∞), and is calculated as follows:
[0107]
[0108] Among them, Tmax This represents the maximum number of iterations.
[0109] Step 3: Perform a random walk on the sample using an improved feedback mechanism. The improved feedback mechanism is adjusted as follows:
[0110]
[0111]
[0112]
[0113] β is the sharing coefficient, and γ is the feedback factor.
[0114] Step 4: Calculate the adaptive switching probability. The improved adaptive switching probability is:
[0115]
[0116] in, Let be the switching probability of the i-th population at the t-th iteration. f i t , These are the optimal fitness value, the i-th population fitness value, and the worst fitness value at the t-th iteration, respectively.
[0117] The population matrix is then iteratively optimized using a butterfly algorithm based on the hybrid gray wolf algorithm. A global search is performed when the random number is less than the adaptive switching probability; otherwise, a local search is performed. The improved search method is as follows:
[0118]
[0119]
[0120]
[0121]
[0122] in, Let g represent the spatial positions of the i-th, j-th, and k-th butterflies in the t-th iteration, where j and k are randomly selected. * f represents the current optimal butterfly population position. i Let X be the fitness value of the i-th butterfly, r be a random number between [0,1], and g1, g2, and g3 be the best, second best, and third best individuals in the current population, respectively. k (i,j) represents the coordinates of the j-th individual (j=1,2,...,N, where N is the population size) in the i-th population (i=1,2,...,d, where d is the spatial dimension), after being guided by the best, second-best, and third-best individuals, respectively.
[0123] If the improved algorithm performs a global search, it first executes formulas (18) to (20), and then executes the first formula of formula (21) to perform a global search. If it performs a local search, it executes the second formula of formula (21) to perform a local search.
[0124] Finally, the new population is subject to boundary crossing handling. Specifically:
[0125]
[0126] Where x1 is the new position of x after processing, and r, r1, and r2 are random numbers between [0,1].
[0127] Step 5: Update the optimal population position and parameter c.
[0128] Next, determine whether the exit condition for the loop has been met based on the current results. If the fitness value reaches the optimal value or the maximum number of iterations is reached, then the optimal population position is the optimized sensor node position; otherwise, repeat steps Step 3 and Step 4.
[0129] like Figure 5 The graph shows a comparison between the improved algorithm HGWABOA of this invention and other existing algorithms in the case of 40 nodes in a 100*100 area. It can be seen from the graph that the improved algorithm has a fast convergence speed and good performance. Figure 6 This is the final node deployment coverage diagram of the improved algorithm of this invention and six comparative algorithms in a 100*100 area with 40 nodes. It can be clearly seen from the diagram that the improved algorithm HGWABOA has the most uniform coverage, with small coverage holes and redundancy.
[0130] Figure 7 Is Figure 3 The results, presented in four scenarios, demonstrate that the improved algorithm HGWABOA effectively reduces node redundancy and energy consumption while maintaining almost constant coverage. For the GWO algorithm, node coverage increases by approximately 4%, redundancy decreases by approximately 3%, and energy consumption decreases by approximately 7%. For the IWOA algorithm, energy consumption decreases by approximately 6% while maintaining coverage and node utilization. The improved algorithm HGWABOA reduces energy consumption by approximately 11%.
[0131] The specific embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited to the above embodiments. Within the scope of knowledge possessed by those skilled in the art, various changes can be made without departing from the spirit of the present invention.
Claims
1. A multi-target wireless sensor network coverage method based on the hybrid gray wolf-adaptive butterfly algorithm, characterized in that: Step 1: Build a model to optimize the multi-target coverage of wireless sensor networks from three aspects: minimum coverage, minimum node redundancy, and minimum energy consumption; Step 2: Initialize the population matrix, calculate the fitness of each population according to the model function, and then calculate its aroma intensity; Step 3: Perform a random walk on the random individual using an improved feedback mechanism. If the fitness of the random individual is better, move it in the direction guided by both the random individual and the best individual; otherwise, move it only in the direction guided by the best individual. Step 4: Calculate the adaptive switching probability, determine the optimization method based on the random value and the magnitude of the adaptive switching probability, optimize the population position using the butterfly algorithm of hybrid gray wolves, and handle the population going out of bounds. Step 5: Update the optimal population position and parameter c. If the fitness value reaches the optimal or the maximum number of iterations is reached, the optimal population position is the optimized sensor node position. Otherwise, repeat Step 3 and Step 4. Step 1 specifically refers to: Step 1.1: Establish the coverage function; Assume the WSN deployment area is of length [length missing]. Width A rectangle, and arrange within the area. The sensing radius of the same specification is The sensor, of which the first Each sensor node is represented as Divide the area into One monitoring point, Record each to be monitored as Set up monitoring points Coordinates are , Sensor coordinates for The overall coverage of the wireless sensor network should be: （4）； Step 1.2: Estimate the maximum number of nodes in the network; During deployment, when the utilization rate of the area covered by a node is greater than 50%, the node should be deployed. To achieve near-complete coverage, the maximum estimated number of nodes to be deployed within the area should be: （5）； （6）； （7）； Step 1.3: Establish the node redundancy rate function; The node redundancy rate is defined as the ratio of the actual coverage area of a sensor to its capable coverage area. Let there be nodes within the region... The sensing radius of the same specification is The sensor node redundancy rate is calculated as follows: （8）； Step 1.4: Establish the energy consumption function and build a multi-objective coverage model; By summing and normalizing the movement distances of the sensor nodes, the energy consumption function is obtained as follows: （9）； The multi-target coverage model is as follows: （10）。 2. The multi-target wireless sensor network coverage method based on the hybrid gray wolf-adaptive butterfly algorithm according to claim 1, characterized in that, Step 2 specifically refers to: Step 2.1: Initialize the population matrix with dimensions pop × 2N, where pop is the population size and N is the number of nodes. The initialization formula is as follows: （11）； in, This is the initial position matrix of all butterfly populations. For butterfly population size, For spatial dimensions, For each element whose value is in the range [0,1] OK A random matrix of columns, , The upper and lower bounds are defined, and each row vector in the matrix represents the position of a set of sensor nodes; Step 2.2: Calculate the fragrance intensity of each butterfly population using the following formula; （12）； in, It is the size of the fragrance that other butterflies can detect. It is the sensory morphology coefficient. It is the stimulus intensity, i.e., the fitness value of the function, which is calculated by substituting the node position of each butterfly population into the cover model. It reflects the degree of absorption and its value ranges from [0,1]. Sensory morphology coefficient The value of is within the range of [0,∞), and its calculation method is as follows: （13）； in, This represents the maximum number of iterations.
3. The multi-target wireless sensor network coverage method based on the hybrid gray wolf-adaptive butterfly algorithm according to claim 1, characterized in that, Step 3 specifically refers to: If the fitness of a random butterfly is better than that of the current butterfly, it will move under the combined guidance of both. Otherwise, it will move towards the optimal butterfly population. The feedback mechanism has been improved by further adjusting its scaling factor. The improved feedback mechanism adjustment method is as follows: （14）； （15）； （16）； In the formula, For the sharing coefficient, It is a feedback factor.
4. The multi-target wireless sensor network coverage method based on the hybrid gray wolf-adaptive butterfly algorithm according to claim 1, characterized in that, Step 4 specifically refers to: Step 4.1: Calculate the adaptive switching probability; The switching probability is adaptively improved so that it determines the search method based on the current iteration. The improved adaptive switching probability is: （17）； in, No. During the nth iteration The switching probabilities corresponding to each population , , The first The optimal fitness value at the nth iteration, the th Each population's fitness value and worst-fit value; Step 4.2: Iteratively optimize the population matrix using the butterfly algorithm of the hybrid gray wolf algorithm; When the random number is less than the adaptive switch probability, a global search is performed; otherwise, a local search is performed. The improved search method is as follows: （18）； （19）； （20）； （21）； in, , , Indicates the first The second iteration , , The spatial location of a butterfly, among which and It was randomly selected. This represents the current optimal location for the butterfly population. For the first The fitness value of a single butterfly. A random number between [0,1] , , These are the best, second best, and third best individuals in the current population, respectively. For the first The first in each population The coordinates of each individual are the position coordinates after being guided by the best, second-best, and third-best individuals, respectively. For population size, Spatial dimension; Step 4.3: Handle the new population from crossing boundaries; Some of the individuals that crossed the boundary were reassigned to the boundary, while the rest were randomly reassigned. Specifically: （22）； in, for The new location after processing It is a random number between [0,1].
5. The multi-target wireless sensor network coverage method based on the hybrid gray wolf-adaptive butterfly algorithm according to claim 1, characterized in that, Step 5 specifically includes: Step 5.1: Update the optimal population position and parameter c; If the updated butterfly population has lower fitness when substituted into the coverage model, then replace the population; otherwise, leave it unchanged. After replacing all populations, select the optimal population for updating. Step 5.2: Determine if the loop exit condition has been met; If the fitness value reaches the optimal value or the maximum number of iterations is reached, then the optimal population position is the optimized sensor node position; otherwise, repeat Step 3 and Step 4.