Smart City Real-Time Safety Management Drones Incorporating Swarm-Based Collaborative Control

Abstract

This invention discloses a smart city real-time safety management drone that incorporates a swarm-based collaborative control method, belonging to the field of drone technology. The technical solution includes a drone body and a swarm control platform. The drone body has a sliding groove, and two connecting strips are slidably connected to the inner side of the sliding groove. This invention improves the original artificial swarm algorithm using a K-means clustering algorithm module. The original swarm algorithm generates the initial population entirely randomly; therefore, the maximum-minimum distance product method is used to disperse the initial population, overcoming its randomness. This reduces blind spots in drone data collection, providing more accurate data support for the swarm-based collaborative control method. This enables rapid real-time video transmission in the event of emergencies or catastrophic violent incidents, allowing commanders to make scientific decisions and judgments, thus improving responsiveness.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) technology, and more specifically, to a smart city real-time safety management UAV that incorporates a swarm-based collaborative control method. Background Technology

[0002] Unmanned aerial vehicles (UAVs) are unmanned aircraft controlled by radio remote control equipment and onboard program control devices, or operated autonomously by an onboard computer, either completely or intermittently. Safe management UAVs can use onboard high-definition cameras to continuously capture images and transmit the information and images back to the ground. In the event of emergencies or catastrophic violent incidents, they can quickly transmit real-time video footage to commanders for scientific decision-making and judgment.

[0003] However, the process of reaching the site involves navigating many complex scenarios, such as densely populated areas or ports. Existing manually controlled drones suffer from the following problems:

[0004] 1. The automatic path planning function of drones cannot meet the actual needs. The path planning algorithm has the characteristics of slow clustering, weak local search ability, and low convergence speed in the later stage of the algorithm.

[0005] 2. The data collection blind spots of drones are large and difficult to adjust, which affects the drone's path planning and may lead to misjudgments by the commander, affecting the ability to respond to emergencies. Summary of the Invention

[0006] To address the problems existing in the prior art, the purpose of this invention is to provide a smart city real-time safety management drone that incorporates a swarm-based collaborative control method.

[0007] To address the aforementioned technical problems, the present invention adopts the following technical solution.

[0008] A smart city real-time safety management drone incorporating swarm-based collaborative control includes a drone body and a swarm control platform. The drone body has a sliding groove with two connecting straps slidably connected to its inner side. A camera is fixedly connected between the two connecting straps. The inner side of the sliding groove is connected to the interior of the drone body. The other ends of the two connecting straps extend into the interior of the drone body. A winding motor is fixedly installed inside the drone body, and a take-up wheel is fixedly connected to the output shaft of the winding motor. One end of each connecting strap inside the drone body is fixedly connected to the take-up wheel. The drone body has two embedding slots located on the left and right sides of the sliding groove, respectively. A magnetic strip is fixedly connected to the inner side of each embedding slot. The camera makes smooth contact with and is magnetically connected to the magnetic strip. A fixed platform is detachably mounted on the bottom of the drone body, and an adjustment mechanism is fixedly installed on the drone body.

[0009] The main body of the drone is divided into hired bee drones and non-hired bee drones. The hired bee drones, non-hired bee drones, and the hive control platform constitute an artificial bee swarm drone system. The hive control platform includes:

[0010] The interactive data processing module is used to process food source information and path information collected by hired and non-hired drones, with the food source information being destination area information;

[0011] The K-means clustering algorithm module is used to perform one K-means iterative clustering on the collected data, and update the honeycomb control platform data with the new data of each cluster according to the cluster division;

[0012] The first signal transceiver module is used to interactively transmit data between the data processing module and the hired bee drones and non-hired bee drones, to obtain collected data and send control commands.

[0013] The data storage module is used to store food source information and path information for later retrieval, viewing, and use;

[0014] Both the hired bee drones and the non-hired bee drones include:

[0015] The data acquisition module is used to collect food source information and route information;

[0016] The second signal transceiver module is used to connect with the first signal transceiver module to send acquired data and receive control commands.

[0017] As a further description of the above technical solution:

[0018] An arc-shaped transition plate is integrally formed at the connection between the sliding groove and the main body of the drone, and the gathering wheel is provided with three gathering grooves.

[0019] As a further description of the above technical solution:

[0020] The connecting strap on the front is a single belt, while the connecting strap on the back consists of two belts, with the ends of the two belts near the camera being integrally formed.

[0021] As a further description of the above technical solution:

[0022] The connecting strap on the front is a single belt, while the connecting strap on the back consists of two belts, with the ends of the two belts near the camera being integrally formed.

[0023] As a further description of the above technical solution:

[0024] The fixed platform includes an equipment mounting platform. The bottom of the drone body is integrally formed with two connecting strips. The left side of the equipment mounting platform is inserted into the left connecting strip, and the right side of the equipment mounting platform is fixedly connected to the right connecting strip by bolts.

[0025] As a further description of the above technical solution:

[0026] The adjustment mechanism includes a threaded post that passes through the drone body and is threadedly connected to it. One end of the threaded post inside the drone body is rotatably connected to a connecting block. The bottom end of the connecting block is rotatably connected to a tensioning wheel, which contacts the connecting belt on the front.

[0027] As a further description of the above technical solution:

[0028] The basic algorithm steps of the K-means clustering algorithm module are as follows:

[0029] Step 1: Set the number of leader bees, follower bees, and scout bees (generally, the number of leader bees should equal the number of follower bees); the maximum number of iterations (MCN) and the control parameter Limit; the current iteration number (Cycle), initialized to 1; the number of clusters (k); initialize the bee colony using the maximum-minimum distance product method, generating {Z1, Z2, ..., Z...} N} initial bee colonies;

[0030] Step 2: Perform a clustering of the initial bee colony, calculate the fitness of each bee, sort them according to fitness, and designate the first half as leader bees and the second half as follower bees.

[0031] Step 3: The leader bee searches its neighborhood to obtain a new position. According to the greedy selection principle, if the fitness of the new position is greater than that of the original position, the original position is updated with the new position; otherwise, the original position is kept unchanged. After all the leader bees have completed the neighborhood search, the probability P is calculated.

[0032] Step 4: The following bees use the calculated probability P and select a leader bee based on the roulette wheel principle. In principle, the larger P is, the greater the fitness value of the leader bee i is, and the greater the probability of being selected by the following bees. After the following bees complete the selection of the leader bee, they search the neighborhood and select the position with high fitness according to the greedy selection principle.

[0033] Step 5: After all the following bees have completed the search, use the obtained positions as cluster centers, perform one K-means iterative clustering on the dataset, and update the bee colony with the new cluster centers of each cluster according to the cluster division.

[0034] Step 6: If a leader bee does not change its result after Limit iterations, it becomes a scout bee and a new position is randomly generated to replace the original position.

[0035] Step 7: If the current iteration count is greater than the maximum number of iterations (MCN), then the iteration ends and the algorithm terminates; otherwise, proceed to step 2, Cycle = Cycle + 1.

[0036] Compared with the prior art, the advantages of this invention are:

[0037] (1) In this scheme, the K-means clustering algorithm module is used to improve the original artificial bee colony algorithm. The initial population of the original bee colony algorithm is generated randomly. Therefore, the maximum minimum distance product method is used to disperse the initial population to overcome the randomness of the initial population. The KMC algorithm divides the samples into multiple classes according to the similarity between the samples. Samples of the same class have the highest possible similarity, and samples of different classes have the lowest possible similarity. This algorithm has the characteristics of fast clustering and strong local search ability, which can effectively solve the problem of low convergence speed in the later stage of the bee colony algorithm.

[0038] (2) In this scheme, after the UAV is launched, the winding motor can be controlled to work according to the flight attitude and environmental factors, driving the winding wheel to rotate, thereby enabling the winding and unwinding of the two connecting belts fixed on the winding wheel. Since the two connecting belts are slidably installed on the inner side of the sliding groove, the camera between the two connecting belts is ensured to surround the UAV body, thereby adjusting the position of the camera on the UAV body, facilitating the collection of video data of the area detected by the UAV body, reducing the blind spots of the UAV collection, and providing more accurate data support for the swarm collaborative control method, so that in the event of emergencies or catastrophic violent events, real-time on-site video images can be quickly transmitted, providing commanders with scientific decision-making and judgment, and improving the ability to respond to emergencies. Attached Figure Description

[0039] Figure 1 This is a schematic diagram of the structure of the present invention;

[0040] Figure 2 This is a side sectional view of the present invention.

[0041] Figure 3 This is a schematic diagram of the connection structure between the winding motor and the take-up wheel of the present invention;

[0042] Figure 4 This is a schematic diagram of the connection structure between the camera mounting base and the roller of the present invention;

[0043] Figure 5 This is a schematic diagram of the structure of the two connecting strips of the present invention;

[0044] Figure 6 This is a schematic diagram of the principle structure of the present invention.

[0045] Explanation of the labels in the diagram:

[0046] 1. Drone body; 101. Hired drone; 102. Non-hired drone; 103. Data acquisition module; 104. Second signal transceiver module; 2. Sliding groove; 3. Connecting belt; 4. Camera; 41. Camera body; 42. Camera mounting base; 43. Roller; 5. Winding motor; 6. Winding wheel; 7. Magnetic strip; 8. Fixed platform; 81. Equipment mounting platform; 9. Adjustment mechanism; 91. Threaded column; 92. Connecting block; 93. Tensioning wheel; 10. Hive control platform; 11. Interactive data processing module; 12. K-means clustering algorithm module; 13. First signal transceiver module; 14. Data storage module; 15. Arc-shaped transition plate; 16. Connecting strip. Detailed Implementation

[0047] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention;

[0048] Please see Figure 1-6 A smart city real-time safety management drone that introduces a swarm-style collaborative control method includes a drone body 1 and a swarm control platform 10. The drone body 1 has a sliding groove 2, and two connecting strips 3 are slidably connected to the inner side of the sliding groove 2. A camera 4 is fixedly connected between the two connecting strips 3. The inner side of the sliding groove 2 is connected to the interior of the drone body 1. The other ends of the two connecting strips 3 extend into the interior of the drone body 1. A winding motor 5 is fixedly installed inside the drone body 1. The output shaft of the winding motor 5 is fixedly connected to a take-up wheel 6. One end of each of the two connecting strips 3 inside the drone body 1 is fixedly connected to the take-up wheel 6. The drone body 1 has two embedding slots, which are located on the left and right sides of the sliding groove 2, respectively. A magnetic strip 7 is fixedly connected to the inner side of the embedding slot. The camera 4 is in smooth contact with the magnetic strip 7 and is magnetically connected. A fixed platform 8 is detachably installed at the bottom of the drone body 1. An adjustment mechanism 9 is fixedly installed on the drone body 1.

[0049] The main body of the drone 1 is divided into hired bee drones 101 and non-hired bee drones 102. The hired bee drones 101, non-hired bee drones 102, and the swarm control platform 10 constitute an artificial bee swarm drone system. The swarm control platform 10 includes:

[0050] The interactive data processing module 11 is used to process the food source information and path information collected by the hired bee drone 101 and the non-hired bee drone 102. The food source information is the destination area information.

[0051] K-means clustering algorithm module 12 is used to perform one K-means iterative clustering on the collected data, and update the data of the honeycomb control platform 10 with the new data of each cluster according to the cluster division.

[0052] The first signal transceiver module 13 is used to interactively transmit data between the data processing module 11 and the hired bee drone 101 and the non-hired bee drone 102, to obtain collected data and send control commands.

[0053] Data storage module 14 is used to store food source information and path information for subsequent retrieval, viewing and use;

[0054] Both the hired bee drone 101 and the non-hired bee drone 102 include:

[0055] Data acquisition module 103 is used to collect food source information and path information;

[0056] The second signal transceiver module 104 is used to connect with the first signal transceiver module 13 to send acquired data and receive control commands.

[0057] In this invention, a K-means clustering algorithm module 12 is embedded into the original bee colony algorithm to perform KMC clustering. This makes the distribution of the hired bee drones 101 and the non-hired bee drones 102 more dispersed after the population update, reducing the impact of randomness on the algorithm and improving the convergence speed in the later stages. When this algorithm is applied to the path planning problem of the drone body 1, it can effectively solve the path planning problem of the drone body 1. During the solution process, the improved algorithm has a faster convergence speed, stronger local search ability in the later stages, and the path obtained is shorter, has a lower total angle, and has a lower cost compared to the original bee colony algorithm. Furthermore, the average quality of the solution obtained in multiple optimizations is better.

[0058] The artificial bee colony algorithm is a new type of intelligent optimization algorithm that simulates the honey-gathering process of bees. It consists of three parts: food source, hired bees, and non-hired bees.

[0059] Food source: The food source is the nectar source. In any optimization problem, the feasible solution is given in a certain form. In the artificial bee colony algorithm, the food source is the feasible solution to the optimization problem and is the basic object to be processed in the artificial bee colony algorithm.

[0060] Hired bee drone 101: Non-hired bee drone 102 corresponds to the location of the leader bee (foraging bee) and the food source. One food source corresponds to one leader bee. In the artificial bee colony algorithm, the number of food sources is equal to the number of leader bees. The task of the leader bee is to discover food source information and share it with the follower bees with a certain probability. The calculation of the probability is the selection strategy in the artificial bee colony algorithm, which is generally calculated by using the roulette wheel method based on the fitness value.

[0061] Non-employed bee drone 102: The non-employed bee drone 102 includes follower bees (observation bees) and scout bees. Follower bees select food sources within the recruitment area of ​​the hive control platform 10 based on the nectar source information provided by the leader bees, while scout bees search for new food sources near the hive. In the artificial bee colony algorithm, follower bees search for new food sources near the food sources based on the information transmitted by the leader bees and make greedy selections. If a food source is not updated after a certain number of times, the leader bee becomes a scout bee, and the scout bee finds a new food source to replace the original food source.

[0062] After the drone is launched, the winding motor 5 can be controlled to work according to the flight attitude and environmental factors, driving the take-up wheel 6 to rotate. This allows the two connecting belts 3 fixed to the take-up wheel 6 to be extended and retracted. Since the two connecting belts 3 are slidably installed inside the sliding groove 2, the camera 4 between the two connecting belts 3 is ensured to surround the drone body 1, thereby adjusting the position of the camera 4 on the drone body 1. This facilitates the collection of video data of the area detected by the drone body 1, reduces blind spots in drone data collection, and provides more accurate data support for swarm-style collaborative control. In the event of emergencies or catastrophic violent incidents, real-time on-site video footage can be quickly transmitted, enabling commanders to make scientific decisions and judgments, and improving responsiveness.

[0063] Please see Figure 2 and Figure 3 Among them, the sliding groove 2 and the main body of the UAV 1 are integrally formed with an arc-shaped transition plate 15, and the gathering wheel 6 is provided with a gathering groove.

[0064] In this invention, the arc-shaped transition plate 15 facilitates local support of the connecting strip 3, avoids excessive wear of the connecting strip 3, and improves the service life of the connecting strip 3.

[0065] Please see Figure 1 and Figure 5Among them, the front connecting strap 3 is a single belt, and the back connecting strap 3 is composed of two belts, with the ends of the two belts near the camera 4 being integrally formed.

[0066] In this invention, since the connecting belt 3 on the front and the connecting belt 3 on the back are divided into single belt and double belt, when the retracting wheel 6 retracts and extends the two connecting belts 3, it can avoid the two connecting belts 3 from interfering with each other. At the same time, by using one retracting wheel 6 to operate the two connecting belts 3, it can ensure the synchronicity of the movement of the two connecting belts 3. The structure is simple and novel and can meet the adjustment work of the camera 4.

[0067] Please see Figure 2 The camera 4 includes a camera body 41, and a camera mounting base 42 is fixedly connected to the top of the camera body 41 by bolts. One end of each of the two connecting straps 3 is fixedly connected to the camera mounting base 42. A roller 43 that is magnetically connected to the magnetic strip 7 is rotatably connected to the camera mounting base 42.

[0068] In this invention, the connecting strap 3 is installed on the camera body 41 through the camera mounting base 42, which facilitates the detachable replacement of the camera body 41 and ensures the connection stability between the camera mounting base 42 and the drone body 1. The magnetic connection between the roller 43 and the magnetic strip 7 increases the contact points, thereby further improving the stability of the camera body 41.

[0069] Please see Figure 2 The fixed platform 8 includes an equipment mounting platform 81. The bottom of the drone body 1 is integrally formed with two connecting strips 16. The left side of the equipment mounting platform 81 is inserted into the left connecting strip 16, and the right side of the equipment mounting platform 81 is fixedly connected to the right connecting strip 16 by bolts.

[0070] In this invention, by installing a detachable device mounting platform 81 between the two connecting bars 16, an additional device mounting position can be conveniently provided at the bottom of the drone body 1, so that other devices such as ultrasonic sensors, GPS positioning, etc. can be installed as needed.

[0071] Please see Figure 1 and Figure 2 The adjustment mechanism 9 includes a threaded post 91, which is inserted into the drone body 1 and threadedly connected to the drone body 1. One end of the threaded post 91 located inside the drone body 1 is rotatably connected to a connecting block 92. The bottom end of the connecting block 92 is rotatably connected to a tension wheel 93, which contacts the connecting belt 3 on the front.

[0072] In this invention, the connecting block 92 is moved up and down by rotating the threaded column 91, thereby controlling the degree of contact between the tensioning wheel 93 and the connecting belt 3 on the front, and adjusting the tension of the connecting belt 3 as needed.

[0073] Please see Figure 6 The basic algorithm steps of the K-means clustering algorithm module 12 are as follows:

[0074] Step 1: Set the number of leader bees, follower bees, and scout bees. Generally, the number of leader bees should equal the number of follower bees. Set the maximum number of iterations (MCN) and the control parameter Limit. Set the current iteration number (Cycle), with an initial value of 1. Set the number of clusters (k). Initialize the bee colony using the maximum-minimum distance product method, generating {Z1, ...

[0075] Z2....Z N} initial bee colonies;

[0076] Step 2: Perform a clustering of the initial bee colony, calculate the fitness of each bee, sort them according to fitness, and designate the first half as leader bees and the second half as follower bees.

[0077] Step 3: The leader bee searches its neighborhood to obtain a new position. According to the greedy selection principle, if the fitness of the new position is greater than that of the original position, the original position is updated with the new position; otherwise, the original position is kept unchanged. After all the leader bees have completed the neighborhood search, the probability P is calculated.

[0078] Step 4: The following bees use the calculated probability P and select a leader bee based on the roulette wheel principle. In principle, the larger P is, the greater the fitness value of the leader bee i is, and the greater the probability of being selected by the following bees. After the following bees complete the selection of the leader bee, they search the neighborhood and select the position with high fitness according to the greedy selection principle.

[0079] Step 5: After all the following bees have completed the search, use the obtained positions as cluster centers, perform one K-means iterative clustering on the dataset, and update the bee colony with the new cluster centers of each cluster according to the cluster division.

[0080] Step 6: If a leader bee does not change its result after Limit iterations, it becomes a scout bee and a new position is randomly generated to replace the original position.

[0081] Step 7: If the current iteration count is greater than the maximum number of iterations (MCN), then the iteration ends and the algorithm terminates; otherwise, proceed to step 2, Cycle = Cycle + 1.

[0082] In this invention, although the artificial bee colony algorithm has advantages such as good global search capability and high robustness, it is prone to premature convergence and low convergence speed in the later stages. To address these issues, the original artificial bee colony algorithm is improved using the K-means clustering algorithm module 12. Since the initial population of the original bee colony algorithm is randomly generated, the minimum distance product method is used to disperse the initial population and overcome its randomness. The KMC algorithm divides samples into multiple classes based on their similarity, with samples in the same class having the highest possible similarity and samples in different classes having the lowest possible similarity. This algorithm features fast clustering and strong local search capability, effectively solving the problem of low convergence speed in the later stages of the bee colony algorithm.

[0083] The above description is merely a preferred embodiment of the present invention; however, the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and its improved concept, should be covered within the scope of protection of the present invention.

Claims

1. A smart city real-time safety management drone that incorporates a swarm-based collaborative control method, comprising a drone body (1) and a swarm control platform (10), characterized in that: The drone body (1) has a sliding groove (2) with two connecting strips (3) slidably connected to the inside of the sliding groove (2). A camera (4) is fixedly connected between the two connecting strips (3). The inside of the sliding groove (2) is connected to the inside of the drone body (1). The other end of the two connecting strips (3) extends into the inside of the drone body (1). A winding motor (5) is fixedly installed inside the drone body (1). A take-up wheel (6) is fixedly connected to the output shaft of the winding motor (5). One end of the two connecting strips (3) inside the drone body (1) is fixedly connected to the take-up wheel (6). The drone body (1) has two embedding slots located on the left and right sides of the sliding groove (2). A magnetic strip (7) is fixedly connected to the inside of the embedding slot. The camera (4) is in smooth contact with the magnetic strip (7) and magnetically connected. A fixed platform (8) is detachably installed at the bottom of the drone body (1). An adjustment mechanism (9) is fixedly installed on the drone body (1). The main body of the drone (1) is divided into hired bee drones (101) and non-hired bee drones (102). The hired bee drones (101), non-hired bee drones (102), and the hive control platform (10) constitute an artificial bee swarm drone system. The hive control platform (10) includes: The interactive data processing module (11) is used to process the food source information and path information collected by the hired bee drone (101) and the non-hired bee drone (102), where the food source information is the destination area information; The K-means clustering algorithm module (12) is used to perform a K-means iterative clustering on the collected data and update the data of the honeycomb control platform (10) with the new data of each cluster according to the cluster division. The first signal transceiver module (13) is used to interactively transmit data between the data processing module (11) and the hired bee drone (101) and the non-hired bee drone (102), to obtain collected data and send control commands; The data storage module (14) is used to store food source information and path information for subsequent retrieval, viewing and use; Both the hired bee drone (101) and the non-hired bee drone (102) include: The data acquisition module (103) is used to collect food source information and path information; The second signal transceiver module (104) is used to connect with the first signal transceiver module (13) to send collected data and receive control commands; The connecting strap (3) on the front is a single strap, while the connecting strap (3) on the back is composed of two straps, and the two straps are integrally formed at the end closest to the camera (4). The fixed platform (8) includes an equipment mounting platform (81). The bottom of the drone body (1) is integrally formed with two connecting strips (16). The left side of the equipment mounting platform (81) is inserted into the left connecting strip (16), and the right side of the equipment mounting platform (81) is fixedly connected to the right connecting strip (16) by bolts. An arc-shaped transition plate (15) is integrally formed at the connection between the sliding groove (2) and the main body of the UAV (1), and the gathering wheel (6) is provided with three gathering grooves. The camera (4) includes a camera body (41), and a camera mounting base (42) is fixedly connected to the top of the camera body (41) by bolts. One end of each of the two connecting straps (3) is fixedly connected to the camera mounting base (42). A roller (43) that is magnetically connected to the magnetic strip (7) is rotatably connected to the camera mounting base (42). The adjustment mechanism (9) includes a threaded post (91), which is inserted into the drone body (1) and threadedly connected to the drone body (1). One end of the threaded post (91) located inside the drone body (1) is rotatably connected to a connecting block (92), and the bottom end of the connecting block (92) is rotatably connected to a tension wheel (93). The tension wheel (93) contacts the connecting belt (3) on the front. The basic algorithm steps of the K-means clustering algorithm module (12) are as follows: Step 1: Set the number of leader bees, follower bees, and scout bees; generally, the number of leader bees should equal the number of follower bees. Set the maximum number of iterations (MCN) and the control parameter Limit. Set the current iteration number (Cycle), with an initial value of 1. Set the number of clusters (k). Initialize the bee colony using the maximum-minimum distance product method, generating {Z1, Z2, ..., Z...} N } initial bee colonies; Step 2: Perform a clustering of the initial bee colony, calculate the fitness of each bee, sort them according to fitness, and designate the first half as leader bees and the second half as follower bees. Step 3: The leader bee searches its neighborhood to obtain a new position. According to the greedy selection principle, if the fitness of the new position is greater than that of the original position, the original position is updated with the new position; otherwise, the original position is kept unchanged. After all the leader bees have completed the neighborhood search, the probability P is calculated. Step 4: The following bees use the calculated probability P and select a leader bee based on the roulette wheel principle. In principle, the larger P is, the greater the fitness value of the leader bee i is, and the greater the probability of being selected by the following bees. After the following bees complete the selection of the leader bee, they search the neighborhood and select the position with high fitness according to the greedy selection principle. Step 5: After all the following bees have completed the search, use the obtained positions as cluster centers, perform one K-means iterative clustering on the dataset, and update the bee colony with the new cluster centers of each cluster according to the cluster division. Step 6: If a leader bee does not change its result after Limit iterations, it becomes a scout bee and a new position is randomly generated to replace the original position. Step 7: If the current iteration count is greater than the maximum number of iterations (MCN), then the iteration ends and the algorithm terminates; otherwise, proceed to step 2, Cycle = Cycle + 1.

Images